Multi-Layered Energy Efficiency in LoRa-WAN Networks A Tutorial

This article has been accepted for publication in a future issue of this journal, but has not been
fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2021.3140107, IEEE Access
Date of publication xxxx 00, 0000, date of current version xxxx 00, 0000.
Digital Object Identifier 10.1109/ACCESS.2017.DOI
Multi-Layered Energy Efficiency in

LoRa-WAN Networks: A Tutorial
IMANE CHEIKH1 , (Student Member, IEEE), RACHID AOUAMI2 , (Member, IEEE), ESSAID
SABIR1,3 , (Senior Member, IEEE), MOHAMED SADIK1 , SÉBASTIEN ROY2 , (Senior Member,
IEEE)
1
NEST Research Group, National Higher School of Electricity and Mechanics, Hassan II University of Casablanca, Morocco
2
Department of Electrical and Computer Engineering, University of Sherbrooke, Québec, Canada
3
Computer Science Department, University of Quebec at Montreal (UQAM), Montreal, QC H2L 2C4, Canada
Corresponding author: Essaid Sabir (e.sabir@ensem.ac.ma).
This paragraph of the first footnote will contain support information, including sponsor and financial support acknowledgment. For
example, “This work was supported in part by the U.S. Department of Commerce under Grant BS123456.”
ABSTRACT Emerging Internet-of-Things (IoT) applications are driving increasing demand for advanced
services in wireless networks, prompting the development of new technologies to address the associated
challenges. Energy efficiency of IoT standards is a key feature targeted by research efforts and industrial
activities, leading to an extensive and growing number of innovative solutions. Low Power Wide Area
Networks (LPWANs) define a class of wireless communication technologies seen as highly relevant for
future IoT development given its long communication range, low-cost devices and interesting energy
management. Long Range Wide Area Network (LoRaWAN) is acknowledged to be the dominant IoT
communication technology. It has allowed broad deployment and unlocked new IoT applications such as
smart cities, asset tracking, etc. This article provides a comprehensive tutorial on the LoRa standard, and
surveys existing solutions, hot topics and future insights for building energy efficient IoT infrastructures and
IoT devices. Indeed, energy efficiency is one of the key factors for successful and sustainable deployments
of IoT applications. More precisely, this article discusses how to meet LoRa/LoRaWAN energy efficiency
across physical layer, medium access control layer, and network layer. Next, extensive pioneering solutions
from related literature are compared and assessed. Finally, insightful conclusions are drawn, and open
problems are listed at the end of this article.
INDEX TERMS LPWANs, LoRaWAN Network, Long-Range, Spreading Factor, Energy Efficiency,
Physical Layer, MAC Layer, NET Layer.
I. INTRODUCTION Type Communications (eMTC) and Ultra-Reliable Low-

A. MOTIVATION AND TRENDS Latency Communications (URLLC). In fact, 5G is designed
NTERNET of Things (IoT) is an emerging ecosystem en- to support Machine-to-Machine (M2M), Device-to-Device
I abling a connected, sustainable and smart environment. It
is making a neverseen progress, with billions of objects being
(D2D) and Device-to-Everything (D2E) communication, the
IoT and the Internet of Vehicles (IoV). Unlike previous
deployed to sense, exchange and share relevant information cellular generations, there are specific provisions in 5G for
about their environment. This emerging digital fabric acts as such applications where low power, small form factor, and
a bridge between physical reality and the Internet. large numbers of connected devices are the norm. In fact, it
It is expected, according to CISCO1 , that the number was a specific intent of the standardization effort leading to
of active connected devices will hit 500 billions by 2030 5G to incorporate IoT-friendly features in cellular standards
empowered by the Fifth Generation (5G) of cellular wireless (e.g., LTE-M, Narrowband-IoT, etc.) [1].
ecosystems. Besides, 5G is leveraged in enhanced Mobile In addition to broadband technologies, IoT development
Broadband communications (eMBB), enhanced Machine- requires two additional network access technologies. On one
hand, high-throughput wireless networks, including Wireless
1 https://www.cisco.com/c/en/us/products/collateral/se/internet-of-things/ Local Area Networks (WLANs) such as WiFi (802.11) and
at-a-glance-c45-731471.pdf cellular systems. This network class is mainly designed to
VOLUME 4, 2016 1
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
I. Cheikh et al.
deliver high data rates and efficient spectrum usage. On the ficiency. When searching for published papers that mention
other hand, once cite wireless sensor networks characterized LoRaWAN energy optimization, academic databases such
by low-cost, ultra low-power devices, as well as low data as IEEE Xplore, MDPI, and other digital libraries show
rates and low duty cycle. This latter access technology fo- significant surge patterns. Recent state-of-the-art surveys on
cuses on energy efficiency and simple transceivers. LoRaWAN [10], [11], [12] [13], to the best of our knowl-
While IoT technology is widely developed, substantial edge, address the challenges and open issues regarding the
research effort remains to be invested to deal with massive physical, MAC layer, and architectural aspects, setting the
connectivity and energy efficiency [2]. The IoT-5G marriage stage for LoRaWAN development. A comprehensive review
sustains implementing sensor-based IoT capabilities into of the energy efficiency of deployed approaches for efficient
robots, actuators and UAVs, to ensure shared coordination network operation, beyond the aspects studied in the existing
and reliable task execution with low latency [3]. Indeed, the literature, is, however, sorely lacking. This paper conducted
incorporation of 5G [4] within IoT systems results in increas- a systematic review of the major search engines in order
ing the transmission rate, reducing transmission delay and to present the existing works on LoRaWAN technology for
guaranteeing network reliability and information security. efficient energy consumption of IoT devices.
IoT enables low-cost objects to access and exchange informa- Table 1 briefly summarizes recent research work on Lo-
tion globally over the Internet [5]. An overview of typical IoT RaWAN networks.
application domains is listed in [6], including smart cities, In this paper, LoRa networks are examined through ex-
healthcare, smart buildings, smart grid and smart industry tensive review of relevant literature to address questions
(industry 4.0). For instance, IoT-enabled home automation regarding their behavior and more particularly their ability
is transforming our daily life, by automatically adjusting the to efficiently optimize energy consumption. The main contri-
temperature according to human presence, and the garage bution aspects of this work are summarized as follows :
door opens automatically upon arrival. In the context of • We provides a comprehensive tutorial covering funda-
smart cities, cohesive communication between autonomous mentals of LoRaWAN network, and surveys the lit-
or semi-autonomous vehicles and infrastructure (e.g., traffic erature dealing with energy efficiency across multiple
lights) lead to improved road safety. Thus, the IoT can both network layers.
enhance quality of life and support global economic activity • We review existing solutions and discuss hot topics
[7]. In concert with this vision, a number of questions must related to energy efficiency at physical (PHY) layer, at
be addressed in the IoT’s continuing evolution [8]: Medium Access Control (MAC) layer and at network
• Can IoT ecosystems ensure connectivity anytime, any-
(NET) layer of LoRa standard.
where and anyhow? • We discuss some of the challenges, open issues, and
• How to ensure continuity of service and interoperability
potential future research directions in optimizing Lo-
between different devices from various manufacturers? RaWAN technology for efficient energy consumption.
• How can energy consumption be reduced to increase the
network lifetime?
C. PAPER ORGANIZATION
• How can a high level of security be provided with low-
cost devices? The remainder of this paper is organized as follows: Section
• How can the rapid evolution of the IoT concept be II deals with LPWAN concept, and compares it to the cellular
effectively managed within massive environments? concept and the most widely known LPWANs technologies.
Low power wide area networks provide to some extent, We survey the main techniques and solutions for energy
many solutions offering low cost technology, low power efficiency at LoRa’s physical, MAC layer and network layer
consumption and satisfactory connectivity to IoT devices. respectively, in Section III, Section IV and Section V re-
Contrary to 4G/5G or WiFi networks, LPWANs in general spectively. Section VI provides insightful discussions and
do not primarily focus on providing high data rates for presents some hot topics and research opportunities. Impor-
IoT devices and/or low latency. Besides, the most desired tant concluding remarks are drawn in Section VII.
features are energy efficiency, connectivity and low cost [9].
By design, 5G mobile network is expected to overcome the II. OVERVIEW OF LPWANS AND LORAWAN
limits of previous cellular standards as a key candidate for A. LORAWAN STANDARD
future IoT enabler. Meanwhile, LPWAN solutions are been LPWAN technologies constitute a valuable complement and
used both to sustain and enable new requirements for IoT- potential alternative to traditional cellular and short-range
critical service cases, notably low cost, low data rate, high wireless technologies for a variety of emerging smart city and
scalability and energy efficiency [1]. machine-to-machine applications [21]. Indeed, the three key
ingredients of LPWANs (high energy efficiency, scalability,
B. RELATED SURVEYS AND OUR MAIN and coverage [22]) are well suited to the development of
CONTRIBUTIONS applications and services for the IoT. Many LPWAN tech-
The academic and industrial communities are becoming in- nologies are currently available which have gained some
creasingly interested in leveraging LoRaWAN’s energy ef- traction in the market, including SigFox, Weightless, Thread,
2 VOLUME 4, 2016
NET/MAC/PHY Energy Efficiency in LoRa-WAN Networks: A Tutorial
Relevants
Year of
Ref. Target Theme Lora Challenges Examined Main Objective OSI
Pub.
Layers
A systematic review of the real-time

UAV-based LoRa (a) The communication range and (b) Packet PHY/
[10] 2021 deployments of the UAV-based LoRa
network loss rate when using UAVs. MAC
communication network.
An overview of the literature

(a) LoRa PHY settings, interference, and
investigating works on LoRa-based
deployment characteristics, (b) LoRaWAN MAC
LoRa PHY and networks, both describing their PHY/
[11] 2021 performance determinants, (c) ADR
LoRaWAN MAC behavior and classifying them MAC
enhancements and alternative solutions, and (d)
according to their technological
Synchronized and time slotted schemes.
advances.
(LPWAN
technologies)
(a) Power consumption, (b) Security, (c) A comparative performance of the PHY/
[12] 2021 LoRaWAN,
Latency, and (d) Throughput perspectives. LPWAN technologies. MAC
NB-IoT and
LTE-M
LoRaWAN A systematic review of the literature

(a) Resource allocation, (b) Network planning, PHY/
[13] 2021 technology results focusing on specific aspects that
and (c) Mobility support. MAC
optimization affect the LoRaWAN performance.
Using the impact of the confirmed traffic: The

A survey on confirmed LoRaWAN
authors studied the impact of (a)
LoRaWAN traffic literature for joint common
[14] 2020 Retransmissions, (b) Confirmed traffic ratios, (c) MAC
Confirmed traffic themes, open challenges, and potential
SFs, (d) Receive windows, and (e)
solutions.
Multi-gateway networks.
Network performance analysis works of LoRa

An overview of LoRa technology is
technology are listed here, and conclusions
LoRa PHY and given, along with relevant work on PHY/
[15] 2020 reached, in terms of : (a) Interference
LoRaWAN MAC network performance in different MAC
Measurements, (b) Multi-hop communications,
scenarios.
and (c) Communication Range.
(a) Energy Efficiency, Success and Coverage

Probability, (B) Channel Throughput of A unified analytical model for
LoRaWAN MAC P-ALOHA, S-ALOHA and non-persistent performance analysis under three basic
[16] 2020 MAC
methods carrier-sense multiple access (NP-CSMA), and random access mechanisms is
(C) Channel access enhancements for: developed.
S-ALOHA, CSMA and Scheduled MAC.
A comprehensive review focusing on

recent solutions for ADR as an
ADR (a) Resource allocation scheme and (b) Energy PHY/
[17] 2020 optimization approach to improve
optimization consumption. MAC
throughput, energy efficiency and
scalability.
A comprehensive review with a

LoRaWAN Mesh
[18] 2020 Multi-hop proposals for LoRaWAN comparative analysis and classification NET
Networks
of multi-hop proposals for LoRaWAN.
No paper classification is done, the authors

review some selected articles about the
LoRa PHY and challenges faced and the proposed solutions in The main challenges and solutions PHY/
[19] 2019
LoRaWAN MAC terms of : (a)Impact of LoRa Imperfect along with the research opportunities. MAC
Orthogonality, (b) Interference Measurements,
and (c) Communication Range.
Authors explore research challenges in terms of : Performance measures, current

LoRa PHY and (a) Energy Consumption, (b) Commununication solutions, key insights and open issues PHY/
[20] 2019
LoRaWAN MAC Range, (c) Multiple Access, (d) Error are addressed based on the research MAC
Correction, and (e) Security. challenges.
Comprehensive tutorial covering

Energy APP/
Energy efficiency across multiple network layers fundamentals of LoRaWAN network,
This Efficiency in NET/
—– including : (a) Resource allocation, (b) MAC and surveys the literature dealing with
work LoRaWAN MAC/
protocols and (c) Multi-hop communications. energy efficiency across multiple
Networks PHY
network layers.
TABLE 1. Summary of recent related surveys.

VOLUME 4, 2016 3
I. Cheikh et al.
NB-IoT, and LoRa, most of them designed to operate in Figure 2 compares the relative strengths of the most popular
subGHz Industrial Scientific Medical (ISM) bands [23]. Fig- LPWAN technologies, thus providing guidance in selecting
ure 1 graphically compares LPWAN technologies with other the best one for a particular scenario. For example, Long
wireless technologies. In particular, the figure highlights the Range (LoRa) provides very long communication range, but
complementarity of LPWAN versus 3G / 4G cellular given no support for real-time data flow [28].
that it features extensive range and coverage at a low cost,
with very low power consumption and a simple ad-hoc infras-
tructure. However, these systems are limited in throughput
and latency. These factors make LPWAN a very promising
candidate for remote monitoring [24].
FIGURE 2. Comparison between LPWANs Technologies [29].
- Narrowband IoT (NB-IoT) : Refers to an LPWAN radio

technology standard developed by the Third Generation Part-
nership Project (3GPP) to support a wide range of cellular
devices and IoT services. Thus, NB-IoT devices with an
average of 200 bytes per day can achieve battery lifetimes of
10 years while covering a distance of 1 km in urban areas and
20 km in rural areas. NB-IoT, unlike LoRaWAN, can support
FIGURE 1. Comparison between LPWAN and cellular technologies [24]. both low latency and high data rates [7].
- Sigfox : It is primarily used to establish IoT networks
In some cases, IOT usage may require highly particular in situations where the volume of data transmitted is low,
functionalities, for which the features of LPWANs are re- according to Sigfox2 an uplink message might have a payload
quired [25]. The main properties of LPWAN technologies of up to 12 bytes and takes about 2 seconds to reach the
required for the successful development of IoT systems are base stations over the air, the range is long, up to tens
shown in Table 2 [26]. of kilometers, and the current consumption is very low,
averaging around mA or 10 mA per transmission. The D-
Main Properties BPSK (Differential Binary Phase-Shift Keying) modulation,
which is part of the UNB (Ultra-Narrow Band) modulation,
Long range of communication is used by SigFox. The most important properties of D-BPSK
modulation are its high efficiency in medium spectrum access
Low power consumption
and ease of implementation, combined with a low bit rate
Low data rate that allows for the use of less expensive components in the
transceiver part [30], [31].
Low cost of device and deployment
- LoRaWAN : In 2015, a group of approximately 500
Simplified network topology and deployment companies formed the LoRa Alliance, resulting in the cre-
ation of a new LPWAN technology standard known as Lo-
Full coverage
RaWAN, which represents the network and MAC stacks that
Network scalability for the capacity upgrade allow network nodes to transmit messages to gateways. LoRa
guarantees secure and reliable communication, as well as
extended battery lifetime and cost savings [32].
TABLE 2. Major properties of LPWAN technologies
The primary goal of this paper is to present an efficient
technology in terms of energy consumption, thus selecting
and focusing on LoRaWAN technology. Authors in [33]
Among LPWAN standards, there are those operating in un-
define it as follows: "LoRaWAN is an energy efficient LP-
licensed bands (e.g., LoRa, SigFox, etc.), and those designed
WAN technology designed to address power consumption
for licensed frequency bands as a component of cellular
and coverage issues in IoT applications".
systems (e.g., NB-IoT, EC-GSM-IoT, LTE-M-IoT, 5G IoT,...)
Table 3 summarizes the main technical specifications of
[27]. The suitability of an LPWAN technology for a specific
LoRaWAN [34].
application depends on requirements in terms of range, data
rate, coverage, power budget, network cost and scalability, 2 https://www.sigfox.com/en/what-sigfox/technology
4 VOLUME 4, 2016
Join Server
Application Server
Network Server
Application Server
LoRaWAN Network Server
FIGURE 3. LoRaWAN architecture.
Standard LoRaWAN centered on chirp spread spectrum modulation (CSS) and

is associated with Forward Error Correction (FEC). There
Spectrum [MHz] Unlicensed
are five configuration elements at the physical layer that
Link Budget 154 dB influence energy consumption, noise impact, and range, in-
cluding: Bandwidth, Central Frequency, Spreading Factor,
Modulation CSS Coding Rate, and Transmission Power [38].
Channel Bandwidth [KHz] 125-500 - MAC Layer : The LoRaWAN is a MAC layer that pro-
vides the media access control mechanism used to connect
Maximum payload [bytes] 243 multiple devices to one or more gateways [39]. Three end-
Power consumption Very low device classes are defined in LoRaWAN: Class A is the most
energy-efficient for low power uplink traffic, Class B is for
Battery life time 2000 mAH 105 months delayed downlink transmissions, and Class C is for end-
Topology Star of stars
devices with non-energy limitations [40].
Security AED CCM 128 B. LORAWAN ARCHITECTURE

LoRaWAN is an open protocol based on the LoRa modu-
Coverage Km ∼5 in urban / ∼20 in rural
lation technology, represented by the physical layer in the
50 kps in Uplink / 290 Bps in OSI model. LoRaWAN technology enables long-distance
Maximum Data rate
Downlink communication in rural (15-20 km) and urban (2-8 km) areas
Transmission technique ALOHA [39]. As shown in Figure 3, the topology of a LoRaWAN
network is a star, with the GW exchanging and forwarding
messages between the ED and the NS, whereby the NS serves
TABLE 3. Technical specifications of LoRaWAN.
as a relay point between applications running on the ED and
the application servers. The ED communicates with the GW
using LoRa, while the GW communicates with the NS using
an IP protocol [41].
Essentially, the LoRaWAN specification includes three pri- - End devices (ED) : They are equipped with sensor nodes
mary elements: the network architecture, the physical layer, that can collect and transmit data to the gateways.
and the MAC layer [35]. - Gateways (GW) : They redirect the received data to the
- Network Architecture : The LoRaWAN network oper- network servers.
ates in a star topology, with end-devices (ED) communicating - LoRaWAN network server : consists of three main
data to gateways (GWs), who then transmit it to a central components: a network server, an application server, and a
network server (NS) via an IP connection [36]. This star join server. The most important functions of the LoRaWAN
topology greatly simplifies operation while also reducing network server are as follows [18]:
network traffic for information routing [37]. • The Network server (NS) : It plays an important
- Physical Layer : The Lora is the physical layer that is role in a LoRaWAN network, filtering duplicated pack-
VOLUME 4, 2016 5
I. Cheikh et al.
ets, providing security, sending acknowledgements, and AWS, to name a few platforms) includes useful functions for
transferring data to the application server [42]. The NS measurement services such as data analysis, long-term stor-
forwards the messages to the Application Server and age, and simple visualization. Cloud services and effective
Join Server. (customer-oriented) dashboards may be the most effective
• The Application server (AS) : The AS forwards all ways to maximize the impact of measurement results on the
packets received from the network server to the specific end-user [47].
associated application. Alternatively, an incoming mes-
sage from an application is forwarded by the application D. LORAWAN PHYSICAL LAYER
server to the network server. 1) Lora Modulation
• The Join Server (JS) : The JS is responsible for the au- At the physical layer, the majority of LPWAN technologies
thentication process of the end devices, both for the gen- employ two modulation techniques: Spread Spectrum (SS)
eration and distribution of the authentication keys. We and Ultra Narrow Band (UNB). In 1997, H.R. Walker pro-
allow two authentication methods on a LoRaWAN net- posed UNB communication and later developed modulation
work: Activation By Personalization (ABP) and Over- methods such as VPSK (Variable Phase Shifting Keying),
The-Air Activation (OTAA). Section E.1 describes in advanced VPSK, VMSK (Very Minimum Sideband Keying),
detail the two authentification methods. and PRK (Phase Reversal Keying) [48]. The UNB modu-
lation technique provides significant bandwidth efficiency.
C. LORAWAN PROTOCOLS STACK Furthermore, it allows signal transmission at a very low
The LoRa protocol stack consists of one mandatory and bandwidth, best suited for small uplink traffic.
two optional classes established at the MAC layer protocol The SS technique, one of the oldest communication tech-
and corresponding to the various possible use cases. LoRa niques, is used in military applications to provide secure
technology covers a series of ISM bands, which vary de- communications, expanding the original signal over a large
pending on the application case and the region in which the frequency band while maintaining the same signal power
LoRaWAN nodes are deployed [43]. In [44], it is stated that [26]. Another feature is the absence of an apparent peak in the
these bands are unlicensed and require conditions in terms spectrum, which merges with noise, making interference and
of maximum transmission power, duty cycle, and bandwidth. interception difficult [21]. Its advantages include interference
Additionally, in [45], they identify maximum duty cycle rejection, multipath suppression, multiple access code divi-
as a major challenge for networks using unlicensed bands, sion, and a high-resolution range. SS transmission refers to a
defining it as "the maximum percentage of time during method in which the signal occupies more bandwidth than is
which an end device can occupy a channel". As a result, required to send the information. Band spreading uses data-
the end device’s emission is limited to these parameters. independent code and synchronized reception at the receiver
LoRaWAN’s duty cycle, depending on the frequency band side for despreading and subsequent data recovery [49].
used, could be 0.1%, 1% , or 10%, with a recommended duty Lora represents the physical layer used in LoRaWAN
cycle of less than 1%. For a value of 1%, the device must networks to achieve low power consumption and long-
wait for 100 times the duration of the previous frame for a range communications via chirp spread spectrum modulation
new transmission in the same channel [46]. (CSS), a SS technique used to modulate the signal using chirp
pulses (variable frequency sinusoidal pulses) [50]. Indeed,
Application the CSS modulation used in LoRa converts each data symbol
into a chirp, defined as a signal whose frequency increases or
LoRa MAC decreases linearly over time. A chirp also refers to a sweep
signal, and each CSS symbol sweeps the bandwidth (BW)
MAC Options once. Whenever a CSS signal’s instantaneous frequency
reaches its highest value, it is transmitted, and the process
Class A Class B Class C begins again at the lowest frequency [51].
LoRa Modulation 2) LR-FHSS

Long Range-Frequency Hopping Spread Spectrum (LR-
Regional ISM Band FHSS) is a new physical layer to address extremely long-
distance and large-scale communication scenarios, such as
FIGURE 4. LoRaWAN protocols stack satellite IoT. Semtech has announced LR-FHSS, a LoRa
physical layer extension. Emerging use cases with increas-
In the current LoRaWAN protocol stack implementation, ingly larger and denser network deployments, such as
the application layer sits on top of the MAC layer, as shown satellite-scale LoRaWAN networks, are driving the exten-
in Figure 4. The LoRaWAN application layer is not defined sion. Its primary aim is to increase network capacity and
at all in the specification. The LoRaWAN application layer robustness by using the Frequency Hopping Spread Spectrum
(for example, running on Loriot, IBM Bluemix, or Amazon (FHSS) modulation technique while maintaining the same
6 VOLUME 4, 2016
communication range as LoRa. LR-FHSS is a fast FHSS by distance. They compute the maximum possible through-
modulation used for uplink communication only; downlink put and the probability of successful transmission, using
communication is achieved with current LoRa because the expressions for perfect and imperfect SF orthogonality [61].
same radios can switch between modulations [52], [53], [54]. The data rate of LoRaWAN varies from 0.3 kb/s to 27 kb/s
depending on the SF used [50].
3) Spreading Factor According to the study [43], by sending a packet with
The spreading factor (SF) is the most relevant variable in a payload of 51 bytes every 10 minutes and assuming a
the CSS system. The CSS order of modulation is given by capacity of 2400 mAh, the battery lifetime for SF7-SF10
M=2SF, indicating that each CSS symbol carries SF bits can reach one year. Comparatively, the battery life for the
[55]. SF refers to the number of chirps required to encode SF12 and SF11 is lower, ranging between 0.5 and 0.8 years.
a bit. For example, SF9 encodes each bit into (29 = 512) Furthermore, it is highlighted in [62] that the ToA, which
chirps [56], providing a flexible trading range for the data refers to the duration of the packet transmission, is dependent
rate [50]. In addition, it improves the spectral efficiency on the value of the SF in use; if the SF increases, the ToA
and capacity of the network. LoRa modulation employs six also increases, implying higher energy consumption. In [63],
orthogonal SFs that influence different data rates. Because the authors estimate the ToA for a standard message with a
these factors are orthogonal different spread signals can be payload of 51 bytes, as illustrated in table 4.
transmitted simultaneously using the same frequency channel
while maintaining communication performance and trading SF ToA
Time on Air (ToA) for communication range [57].
The SF is defined as : 7 112.90
8 195.07
SF = log2 (Rc /Rs ), (1)
9 349.18
where:
- Rc is the chirp rate, 10 616.45
- Rs is the symbol rate. 11 1150.98
The SF can have six different values, ranging from SF=7
to SF=12 [40]. Then there is a compromise between the 12 2138.11
SF and the coverage range. The higher the SF, the larger
the communication range is. As stated above, the SFs are TABLE 4. ToA for a standard message (51 bytes) per (ms).
orthogonal, which means the gateway can receive multiple
transmissions simultaneously on different SFs [58]. The au-
thors of [59] investigate the effects of imperfect orthogonality The actual ToA for a packet in a LoRa network can be
between multiple LoRa SF transmissions. The authors claim defined as follows [64]:
that a LoRa transmission can be interfered with by another SF
transmission if the power of the interfering signal is signifi- ToA = Tpreamble + Tpayload , (2)
cantly greater than the power of the reference signal. Indeed,
the results show that this power difference is about 16 dB. where Tpreamble is the preamble duration and Tpayload is the
This power difference may occur when an interfering signal payload duration. Tpreamble can be calculated as follows:
is close to a receiver or several interfering signals are received
simultaneously. Another research study addresses the effect Tpreamble = (Lpreamble + 4.25)Ts , (3)
of interference on a LoRa network due to transmissions
running concurrently while using the same SF. The results where Lpreamble is the preamble length and Ts represents the
show that transmitting using different SFs may significantly time of one symbol and :
impact LoRaWANs with high density [60].
Another study proposes a detailed analysis of the available Ts = 1/Rs. (4)
uplink throughput on a LoRa network, considering the effects
of co and inter-SF interferences. Two types of interference Based on the above formulas, we can conclude that the
are involved, the first type considering that SFs are not ToA of the LoRa packet gets directly influenced by the SF
perfectly orthogonal. Then the transmission of packets using hence the choice of SF is a decisive factor to ensure proper
different SFs is subject to collisions in the form of inter- network management.
SF interference. Alternatively, if the same SF uses the same
channel, the second type of collision occurs called co-SF 4) Lora Physical Layer Frame
interference. There are two different types of SF allocations: Figure 5 depicts the physical, MAC, and application layer
the SF-random type, in which the SFs uniformly distributed, Lora frame structure [65]. The following fields are included
and the SF-distance type, in which the SFs are distributed in the LoRa packet [64]:
VOLUME 4, 2016 7
I. Cheikh et al.
• Preamble field: Serving for synchronization purposes, way serves many devices and relays messages to a central
comprising eight successive reference chirps, that indi- server [66]. LoRaWAN defines the network protocol for
cates the packet modulation scheme, modulated with the LoRa-based devices [67]. Figure 6 depicts the use of two
same spreading factor as the rest of the packet; distinct keys in message transmission :
• Header field: Two operating modes are available. The - The network session Key : It is used to encrypt the
number of bytes in the header field indicates the FEC frame and ensure that the message gets sent correctly and in
code rate, the length of the payload, and the presence its entirety for better communication. This key gets shared
of CRC in the frame in the default explicit operational between the end device and the network server.
mode; in the second implicit mode of operation, it spec- - The application session Key : It is used to encrypt and
ifies that the code rate and payload in a frame remain decrypt the payload in the frame (application data messages)
fixed. This field is not included in the frame when using and to ensure security. This key gets shared between the end
this mode, which helps to reduce transmission time. A device and the application server [68].
2-byte CRC field is also present, allowing the receiver
to reject packets with invalid headers. The header field,
JOIN REQUEST LORAWAN (JOIN)
including the CRC field, is 4 bytes long and has a coding END DEVICES
SERVER
rate of 1/2. However, the coding rate for the rest of the
frame resides in the PHY header. Note that the length IF OK
of the payload gets determined by the first byte of the

JOIN ACCEPT
header field;
• Payload field: The payload’s maximum size ranges
DERIVES
from 2 to 255 bytes. This field also includes the fol- SESSION KEYS
lowing elements: MAC header (specifies the frame

type, protocol version, and direction); MAC Payload APPSKEY SECURE COMMUNICATION
+
(including actual data); MIC (corresponds to the digital NWKSKEY
signature of the payload);

• CRC Field: It is optional and contains Cyclic Redun-
dancy Check (CRC) bytes to protect the payload from FIGURE 6. OTAA Procedure.
errors (2 bytes).
1) MAC messages
PHY PHY PHY
Preamble
Header Header Payload
CRC Different message types are present in the LoRaWAN MAC
CRC
Layer, and two joining procedures, OTAA and ABP, are
included. One of these procedures must be executed by the
MAC MAC
end devices. In [27], it is stated that in order to join the
MIC
Header Payload LoRaWAN network and exchange data, the end device must
perform an Over-The-Air Activation (OTAA) procedure,
composed of two MAC messages exchanged between the
end device and the network server (Join Request and Join
Frame Frame
Frame Accept), and if the end device loses connection, it must
Header Payload
Port
repeat this procedure. Figure 6 shows the two MAC messages
exchanged during an OTAA procedure.
In [67], the authors present the second type of joining
Device Frame Frame Frame
Address Control Counter Options procedure, Activation by Personalization (ABP), a proce-
dure that allows end devices to be directly connected to
FIGURE 5. LoRa frame structure.
the network without the need for a join-request and a join
accept. However, the first procedure used to join the network
In [27], the authors highlight the single difference between (OTAA) is qualified as the most appropriate and secure way
the structure of uplink and downlink messages: the downlink for an end device to join the network. Once the connection is
messages do not contain the CRC field, as messages should established, the end device starts sending and receiving data
be as short as possible to minimize the effect of any duty messages.
cycle limitations. The messages received by the gateway can be the subject
of an ACK (acknowledgment) by setting the ACK bit if
E. LORAWAN MAC LAYER requested or not, which means that in the first case, the
As mentioned in the previous section, the LoRaWAN MAC gateway sends an ACK to confirm the reception of the
(medium access control) layer allows communication be- data messages, while in the second case, no confirmation is
tween multiple end devices and their gateway (s). The gate- required [69]. Furthermore, in [70], they indicate that for the
8 VOLUME 4, 2016
first use case, two ACKs are received. The first ACK is sent This protocol poses a challenge when the number of
on the same channel where the data was transmitted and T1 nodes is growing considerably. Without a listen before talk
after the reception of the frame, the second ACK is sent on strategy, after each message collision, these messages are
the downlink channel after a timeout of T2 = T1 + 1s, and retransmitted later. In the occurrence of collisions and the
retransmission is performed if the end devices do not receive necessity to retransmit lost packets, the channel capacity will
an ACK. In some cases, gateways are required to send an be decreased [77]. According to the authors of [78], the
ACK as confirmation of the connection between them and throughput of P-ALOHA channels can be used up to 18%
the nodes. If multiple gateways receive the same message, of the total channel capacity. To overcome the challenges of
one of them must respond by sending an ACK to the node in P-ALOHA, the Slotted-ALOHA (S-ALOHA) protocol was
question [71]. introduced. A widely used MAC protocol in local wireless
communications, where the channel time is divided into slots
2) MAC commands and end devices can only send packets at the beginning of
MAC commands are control information sent by the gateway a slot; if two or more nodes send their packets at the same
to the nodes. Each MAC command should have the appropri- time, a collision occurs, indicating that the data is not being
ate response from the recipient nodes. Table 4 lists the MAC sent correctly; otherwise, the data is correctly transmitted and
commands supported by LoRaWAN [72], [73], [74]. no collision occurs. It was mentioned that the throughput of
S-ALOHA channels can use up to 37% of the total channel
Command Description capacity.
In the context of extremely rich and valuable studies
Requests to the end-device a modification of that employ the S-ALOHA protocol, the authors of [79]
LinkADR the data-rate, transmit power, repetition rate or explore the distributed choice of retransmission probabilities
channel.
in Slotted Aloha from a game theoretical perspective. Using
LinkCheckReq
Sent by the end-device to validate the quality a Markov chain analysis, they were able to obtain optimal
of the link. and equilibrium retransmission probabilities and throughput,
Provides an estimation of the received signal before evaluating the impact of adding retransmission costs.
LinkCheckAns power as an indicator of the reception quality In parallel, the authors of [80] analyze the performance of the
of the end-device. Slotted-Aloha-based uplink of a cellular system according to
Allows to set the maximum duty-cycle of a several power differentiation schemes and at another level,
DutyCycle
device for transmission. they deduce the expected throughput and delay in order to
optimize and provide a stability analysis that will serve as
Used to change the reception parameters of the
RXParamSetup
device. an alternative study. Authors of [81] focus on the retransmis-
sion probabilities in LoRaWAN networks using the Slotted
Used by the network server to reset the status Aloha protocol with the goal of maximizing throughput while
DevStatus
of the device.
taking into account that the SFs are perfectly orthogonal.
Allows to modify the definition of the radio They have demonstrated that it is possible to achieve both
NewChannel
channel parameters. satisfactory throughput and limited delays by fine-tuning the
retransmission probabilities and correctly setting the MAC
TABLE 5. LoRaWAN MAC commands. parameters. The authors of [82] consider a model for in-
creasing the average system packet success probability (PSP)
under Pure Aloha. To achieve this, they proposed an opti-
mization model aiming to maximize the average PSP of the
3) MAC protocols system, via a sub-optimal SF allocation method that takes
It is clearly stated in [73], that energy-efficient use is primar- into account the effects of interference using the same or
ily influenced by MAC protocols, designed to manage uplink different SFs.
and downlink messages, node mobility, and determining
network scalability. F. LORAWAN NET LAYER
LoRaWAN uses Pure-ALOHA (P-ALOHA), which allows LoRaWAN employs a single-hop routing model, in which
node senders to select a SF, and a transmission power for gateways in the center communicate and transmit messages
their messages. While the Listen before Talk (LBT) option from terminal devices to the network server. The gateways
is not excluded in LoRaWAN. Even though it is generally are linked to the Internet network via Wifi, 4G, or Ethernet,
an inefficient protocol, pure ALOHA is still widely used allowing data to be sent to the network server via IP. It is
because it offers multiple significant advantages, such as possible to bring together different end-devices and gate-
variable packet size, flexibility in starting transmission, and ways in the same geographical area by isolating frequency
no requirement for time synchronization. As a result, most communications as well as virtual channels within the same
(if not all) LPWAN technologies use pure ALOHA as their frequency channel, keeping in mind that transmissions with
primary MAC protocol [75], [76]. different SFs are orthogonal due to the spread of the spectrum
VOLUME 4, 2016 9
I. Cheikh et al.
Smart City Smart Agriculture
Smart Transport Smart Home
Smart Utilities Smart Grid
Urban Areas Rural Areas

Deployment LoRaWAN Deployment
Opportunities Opportunities
Smart Health
Object Tracking
Smart Broadcasting
Connected Cars Smart Education Smart Manufacture
FIGURE 7. LoRaWAN deployment opportunities.
[83]. H. LORAWAN SERVICES

The single-hop model has several limitations, including In the LoRaWAN network, each device has different classes
the fact that gateways are both complex and expensive to that define its capabilities. In fact, three classes are defined
maintain, particularly for a wide range of communications by the LoRaWAN Alliance and create a trade-off between
[84]: network down-link communication latency and battery-life.
• Several gateways must be installed throughout the area. Figure 7 depicts the three LoRaWAN network classes.
• The existing GW purchases and installs a communica-
tion module for internet use.
• A monthly fee will be charged if the gateway connects to Transmit RX 1 RX 2
the Internet via a private telecommunications network. RX Delay 1
• Several issues of rising costs associated with the com- RX Delay 2
munication module and its monthly fee. Class A

Energy Efficiency
They are also required to listen to all channels at the same
Latency
time and to be constantly connected to the Internet [85]. BCN PNG Transmit RX RX 1 RX 2 BCN
Delay 1
RX Delay 2
G. LORAWAN DEVICES AND APPLICATIONS FIELDS Beacon Period
LoRa networks have been widely deployed for a variety of Class B

applications and research systems. LoRa’s openness makes it
an extremely suitable choice for various IoT deployments. Transmit RX
Common IoT applications include smart buildings, smart
cities, smart agriculture, smart metering and water quality RX Delay 2
measurement [20] [12]. Furthermore, paper [86] claims that Class C

the acquisition costs of LoRa-enabled devices, as well as
FIGURE 8. LoRaWAN end devices classes.
mobile devices and gateways, are low.
In [38], a performance study and analysis of the capabil-
ities of a currently available LoRa transceiver is performed, The three classes can be used together in the same network,
followed by a description of the transceiver’s characteristics and end nodes can switch between them [87]. The three
and a demonstration of its efficient use in an extended appli- classes are as follows:
cation scenario. Certain features, such as simultaneous non- • Class A "All" : This is the class supported by all
destructive transmissions and carrier detection, have been devices in the LoRaWAN network and the most energy-
demonstrated to be useful. The demonstration clearly shows efficient.
that six LoRa nodes can form a network covering 1:5:ha in a It can optimize battery life for years. Aloha-type access
built-up environment, with a potential lifetime of two years to the channels is used. Then, two down-link windows
on two AA batteries, providing data in five seconds and 80% (RX1 and RX2) are available after every up-link com-
reliability. munication; otherwise, the end devices are in sleep
Figure 7 identifies LoRaWAN deployment opportunities in mode; the device can only receive a message from the
both urban and rural areas. gateway during these two windows.
10 VOLUME 4, 2016
LoRaWAN
Challenges
Energy Communication
Multiple Access Error Correction Security
Consumption Range
Link Resource Channel Interference Key Third-Party

Allocation Key Update
Coordination Coding Cancellation Generation Authorization
FIGURE 9. LoRaWAN challenges.
• Class B "Beacon" : Similarly to Class A, while end • Multiple Access : The goal of LoRa networking is
devices receive messages in time-synchronized sched- to connect thousands of end-devices to the network
uled receive slots, the gateway transmits a beacon at while communicating over a limited region and spec-
periodic intervals to synchronize all end devices on the trum. Depending on the application, the possibilities for
network. When a terminal device receives a beacon, it these end-devices transmitting data concurrently vary.
can open a short reception window called a "Ping Slot" Multiple access issues involve two distinct aspects: link
in a previewed form during a regular time slot. coordination and resource allocation.
In terms of latency and power consumption, Class B • Error Correction : LoRa technology is used to transmit
offers a balanced solution [88]. data over long distances. While the message is being
• Class C "Continuous" : For this class, the receive transmitted over the air, the data may become corrupted
window is always open, it is only closed when the end or lost due to channel effects, environmental conditions,
device is transmitting. It is used for real-time applica- or collisions. Current solutions fall into two categories:
tions, so the energy consumption is higher. Among the channel coding and interference cancellation.
three classes, Class C has the lowest latency for the ED • Security : Security is a major concern in all computer
[89]. communications. Eavesdropping, selective forwarding,
and node impersonation are all examples of security
I. LORAWAN CHALLENGES attacks.
The authors of [20] discuss the research challenges of LoRa Figure 8 summarizes the challenges associated with the
networking and divide them into five major components: LoRaWAN network.
• Energy Consumption : The most significant feature
As previously stated, energy efficiency is the most impor-
of LPWAN is its high energy efficiency. This becomes tant feature of LPWAN technologies, thus the idea of evalu-
an important parameter in extending the life of end- ating the energy efficiency of the LoRaWAN network. Unlike
devices. LoRa networks are expected to operate for 5- previous surveys on LoRa technology, this work focuses on
10 years with minimal maintenance. As a result, power the evaluation of energy consumption at the different layers
consumption becomes a significant challenge for LoRa of the OSI model.
networking.
• Communication Range : A long communication range
III. ENERGY EFFICIENCY AT PHYSICAL LAYER
is also an essential component of LoRa technology.
Current LoRa technology is based on chirp spread spec- LoRa is the physical layer used in Lorawan networks to
trum, which is less susceptible to interference. LoRa achieve low energy consumption and long-range communi-
networking will be used in a variety of settings, includ- cation, due to its very appealing advantages. LoRa PHY layer
ing homes, hospitals, schools, and forests. End-devices is gaining popularity among researchers.
will be placed in locations that are open to the air. With In this section, we examine the energy consumption of the
such diverse deployment conditions, signal attenuation, most relevant research works from both the network (config-
propagation losses, and fading must be mitigated in or- urable radio parameters) and the device (resource allocation)
der to improve signal penetration and thus the coverage perspectives.
of LoRa networks [90], [91].
VOLUME 4, 2016 11
I. Cheikh et al.
A. NETWORK SIDE The following is a collection of recent research on the above-

According to the authors of [92], [93], five parameters at the mentioned parameters:
physical layer effect energy consumption and range : In [94], the impact of parameter selection on power con-
sumption and communication reliability is detailed, and an
• Bandwith (BW) : It is the amplitude value related to
algorithm capable of rapidly identifying a good transmission
the frequency domain for the channel used; a higher BW
parameter setting is proposed. The challenge here was to
results in a higher data rate and, as a result, the shortest
determine the setting that minimizes the cost of transmission
transmission time. A higher BW, on the other hand,
energy while also satisfying the required communication
means lower sensitivity since more noise is included.
performance, thus creating a balance between network per-
Most LoRa networks use a channel with a bandwidth of
formance and energy consumption. This work is a first step
125 kHz, 250 kHz, or 500 kHz.
toward developing an automated mechanism for selecting
• Central Frequency (CF) : It is spread over different
LoRa transmission parameters.
frequency channels by exploiting the implementation of
The authors of [96] create an energy model to estimate
pseudo-random channel hopping. CF values depend on
and optimize wireless sensor energy consumption. Based on
local frequency regulations, while LoRaWAN gateways
LoRaWAN Class A, the proposed model integrates the mod-
typically support eight channels, whereas IoT devices
eling of sensor node units, specifically the processing and
usually support at least 16 channels. Specifically, CF
sensor units, using a true IoT application. Furthermore, many
can be configured in the range of 137 to 1020 MHz,
different LoRaWAN transmission modes are investigated in
depending on the legislation in the geographic region
order to select the best mode that can optimize energy con-
[94].
sumption. Moreover, a comprehensive optimization study of
• Spreading Factor (SF) : The LoRa uses six different
LoRaWAN parameters such as SF, CR, BW, communication
programmable SFs, ranging from 7 to 12. This allows up
range, and TP is discussed with the goal of maximizing
to six nodes to transmit on the same channel at the same
sensor lifetime. In [56], an analysis of the performance of
time. In [95], it is noted that increasing the SF value
the LoRaWAN network is presented, including discussion of
without changing the transmission power causes the
the impact of various spreading factors and the power used
energy consumption to increase more rapidly. Higher SF
in transmission frames. The experiment was carried out by
values result in a lower bit rate, but higher sensitivity.
sending packets with the same payload at various spreading
Overall data rates in use are determined by regional
factors and power settings; the results show that using higher
specifications. Table 5 shows the possible data rate (DR)
spreading factors and increasing transmission power reduces
for the EU 863-780 MHz ISM Band [65].
observed packet loss.
Channel, PHY bit RF sen-
In [97], the authors assess the energy consumption of
Data Coding the sensor nodes, taking into account three factors that in-
SF width rate sitivity
Rate Rate
(KHz) (bps) (dBm) fluence energy consumption : transmission distance, data
0 12 125 4/6 250 -137 transfer/collection, and stability. The proposed simulation
1 11 125 4/6 440 -136
system considers three nodes with a LoRa base station, and
the duration of the communication test is approximately
2 10 125 4/5 980 -134
seven days. Three communication parameters, SF, CR, and
3 9 125 4/5 1760 -131 distance, are then varied and regulated; once one of these
4 8 125 4/5 3125 -128 parameters is adjusted, the other variable remains constant.
5 7 125 4/5 5470 -125 The results showed that when the SF is high, the power con-
6 7 250 4/5 11000 -122 sumption is also high, whereas when the CR is low, the power
consumption is low, when performing a power consumption
TABLE 6. Data Rates in EU 863-780 MHz ISM Band and node transmission data measurement according to the
LoRa protocol.
• Coding Rate (CR) : It represents the redundancy used LoRa Backscatter is a CSS modulated system with data
in FEC codes to provide interference protection. A transmission distances of up to 2.8 km and a power consump-
lower CR provides better protection, but it increases tion of only 9.25 watts at a data rate of 37.5 Kbps. When
transmission time, as indicated by : compared to standard LoRa technology, this system reduces
power consumption by nearly 1000 times. The solar panels
CR = 4/(4 + p), with p ∈ {1, 2, 3, 4}. (5) mounted on the passive RF chips can be used to power them
[98].
• Transmission Power (TP): Power control options are The authors of [99] present a PLoRa (Passive Long-Range)
limited, with only six values available: 2, 5, 8, 11, 14, communication technology that provides long-range connec-
and 20 dBm. For the LoRa system deployed in the EU tivity for IoT devices based on ambient excitation, removing
863-870 band, the maximum TP value for use is 14 dBm the need for any dedicated excitation source. The PLoRa tag
in accordance with ETSI regulations. is battery-free and generates energy from radio signals and
12 VOLUME 4, 2016
ambient light, allowing it to communicate with active LoRa of the ADR mechanism is assigned to the network server,
nodes and gateways over long distances in three distinct with the goal of keeping the nodes as simple as possible. The
modes. The experimental results show that the prototype ADR algorithm performed on the node is specified by the
PLoRa PCB tag is able to backscatter an ambient LoRa LoRa Alliance, while the algorithm on the network server is
transmission transmitted from a closely situated LoRa node determined by the network operator. The ADR algorithm on
(20 cm) to a gateway as far as 1.1 km away, delivering the network server can decrease the SF and change the TP, the
284 bytes of data every 24 minutes indoors, or every 17 algorithm on the node can only increase the SF. The part of
minutes outdoors. However, the detection range of PLoRa the ADR performed at the node (ADR-NODE) is primarily
packets is limited to 50 m, so there is a large area of potential intended to increase the SF (and thus reduce the data rate)
improvement. if the uplink transmissions cannot reach the gateway, if a
Monitoring these physical layer parameters is an important downlink frame is not received within a configurable number
factor in ensuring an efficient and controlled use of energy in of frames, the node increases the SF of the next uplink frame.
a LoRaWAN network [100]. This increases the transmission range and therefore the prob-
ability of reaching a gateway. To improve the LoRaWAN
B. SENSOR/IOT DEVICE SIDE network, alternative SF allocation techniques are available,
1) Resource allocation such as:
We present a comprehensive analysis of the most signifi- In [72], the authors propose two algorithms, EXPLoRa-
cant results of the state-of-the-art on energy efficiency of SF and EXPLoRa-AT, to outperform the basic ADR ap-
allocation mechanisms. We classify existing works into two proach. The first algorithm has the advantage of allowing
types of approaches: (a) Single parameter allocation and (b) users to choose SF based on the total number of connected
Multiple parameter allocation, as described below : devices. The second algorithm aims to use an innovative
"ordered water filling" approach that allows the spreading
a: Single parameter allocation factors to be distributed in order to balance the ToA of the
In a LoRaWAN network, a node can not predict the distance packets transmitted by the system’s end devices based on
between it and a gateway at first, but it can estimate the groups of spreading factors. The results show that using this
distance by observing the power signal received from a last technique is very efficient, especially under high load
downlink transmission. If the received signal power of a conditions when the system supports thousands of nodes or
downlink transmission is very high, it can reduce the SF high message rates. But there is no evaluation of the energy
of its next transmission to save power. This SF allocation consumption of the proposed algorithms.
scheme is known as the lowest possible SF allocation scheme The authors of [105] propose a new method for an effi-
[101]. It is important to note that EEC techniques are used cient allocation of the SF in a LoRaWAN network designed
in the lora physical layer to ensure noise and interference specifically to improve network scalability. Compared to the
resistance and to increase receiver sensitivity [66]. On this conventional method, the new method increases network
basis, the LoRaWAN specification includes an Adaptive Data scalability and improves data delivery probability while in-
Rate (ADR) mechanism that allows a network server to select creasing average power consumption by 1 to 8%. In [101], an
both the data rate and the transmission power provided to open source discrete event simulator, is presented to analyze
each node. Furthermore, the authors of [102] state that the the performance of the LoRaWAN network and to investigate
ADR mechanism serves two critical purposes: different schemes for SF allocation. A machine learning
• Increase the global network capacity, solution called the intelligent Decision Tree Classifier (DTC)
• Extend the battery life of the nodes. scheme and intelligent Support Vector Machine (SVM) are
The transmission power is dynamically assigned to a node described. The results show that the proposed systems can
based on its distance from the gateway, thus increasing bat- improve the overall performance of LoRaWAN networks, but
tery life. The authors of [103] state that the ADR mechanism the nodes use their maximum transmission power, which is
reduces energy consumption by five times per payload byte. an area for future improvement of the proposed schemes in
ADR is mentioned by authors of [104] as being used for order to reduce the nodes’ energy consumption.
both energy savings and extending communication range. The authors of [106] suggested a lightweight learning
ADR is accomplished by employing orthogonal different approach suited to the communication parameters of IoT
SFs and varying transmission power. These orthogonal SFs devices and reaching energy efficiency and reliability goals.
allow multiple LoRaWAN end devices to operate on the same In this regard, this approach assigns SFs based on the distance
frequency channel at the same time. When the transmission between the IoT device and the gateway. A new proposed
power is consistently higher than the sensitivity, the ADR al- algorithm to improve the energy efficiency of the ADR
gorithm has the ability to increase the SF whose SNR margin scheme is presented in [107], a comparison between the
is too low, or even decrease it. This algorithm is required for two algorithms is performed, showing that the new pro-
ED movement during the NS SF selection process. posed algorithm is superior in terms of energy consumption,
The ADR scheme [100] operates asynchronously at the with many cases showing more than 100% efficiency im-
LoRa node and the network server. Most of the complexity provement. Paper [109] improved the ADR mechanism by
VOLUME 4, 2016 13
I. Cheikh et al.
Adjusted Allocation Energy

Ref. Advantages Drawbacks Energy Consumption
Parameter Scheme Efficiency
Increases by 1% for a network

Maximizes the correct packet The average power
using two SFs, while for six
SF —— delivery rate (PDR), improves the consumption is raised ⋆
[105] SFs the average consumption
scalability of the network. by 1 to 8%.
is increased by 8%.
A solution based on machine It is always assumed

Intelligent
learning increase the performance that the nodes use the Both schemes consume
DTC and
SF of LoRaWAN network, good maximum transmit slightly more energy, ⋆⋆
[101] SVM
simulation results in terms of power for uplink approximately a 1% increase.
schemes
packet delivery rate (PDR). transmission.
Running at the node, using With a highly loaded

orthogonal different SFs; Increase network, the number Highly efficient in terms of
SF ADR-NODE ⋆ ⋆ ⋆⋆
[100] the global network capacity; of collisions is energy consumption.
Extend the battery life of the nodes. increasing.
A learning solution to adapt device The proposed scheme reduces

Light-weight
communication parameters to the energy consumption per data
distributed
SF environment in order to maximize transfer by more than 50%, ⋆ ⋆ ⋆⋆
[106] learning
energy efficiency and data while increasing the
approach
transmission reliability. probability of success by 30%.
Totally ignores the

slow convergence The energy efficiency is
A new algorithm proposed to
time of the ADR, approximately doubled
SF —— improve the energy efficiency of ⋆ ⋆ ⋆⋆
[107] which is impacted by compared to the basic
ADR.
the variable channel algorithm.
conditions [108].
E-ADR is not suitable

Improves the ADR mechanism by
for applications The gain on energy
taking into account the position of
where some nodes consumption is increased by
mobile devices and their
follow approximately 60.23%, due to
trajectories in order to have a
SF E-ADR random/unknown the reduction of waiting ⋆ ⋆ ⋆⋆
[109] dynamic allocation, and minimize
trajectories because it packets (3 instead of 20) and
the transmission time and energy
was designed for its ability to re-adapt the
consumption as well as the packet
known mobility configuration more quickly.
loss for mobile devices.
patterns.
The authors present the total

energy consumed over the
There has been no number of received packets, as
An E-ADR extension to include
VHMM- investigation into the well as the impact of the
unknown random mobility pattern,
SF based scalability of the retransmission threshold on ⋆ ⋆ ⋆⋆
[110] extremely efficient in terms of PLR
E-ADR proposed mechanism energy consumption; for both
and energy consumption.
in a large network. results, the proposed solution
has a very low energy
consumption.
(⋆): Very Low; (⋆⋆): Low ; (⋆ ⋆ ⋆): High ; (⋆ ⋆ ⋆⋆): Very High.
TABLE 7. Energy efficiency of SFs allocation methods using one configurable parameter.
proposing a new allocation scheme called Enhanced-ADR E-ADR to deal with unknown mobility patterns. This E-
(E-ADR) that can perform dynamic allocation procedures, ADR extension, known as VHMM-based E-ADR, predicts
thus optimizing network transmission time, reducing energy the node trajectory using a Variable order Hidden Markov
consumption, and decreasing overall packet loss. The per- Model (VHMM). It was built on the Waspmote SX1272
formance of E-ADR was evaluated with Waspmote-SX1272 hardware platform. The experimental results show that it is
devices and gateways using several mobility models in a very efficient in terms of packet loss rate (PLR) and energy
smart farm scenario. The results demonstrate that E-ADR consumption.
can reduce, and in some situations, eliminate packet loss Table 7 lists some of the SFs allocation strategies that
and support mobility procedures; In addition, the gain on have been used to improve the performance of LoRaWAN
energy consumption is increased by approximately 60.23%. networks and their energy consumption compared with the
The same authors in [110] propose an expanded version of standard solution using single configurable parameter.
14 VOLUME 4, 2016
b: Multiple parameter allocation their utilities. A channel assignment algorithm was proposed
based on distributing the channel access decision site to
The part of ADR executed on the network server, called
users. Consequently, the LoRaWAN gateways achieved the
ADR-NET, where the algorithm allows the network server
optimal TP for IoT devices when sharing the same CF with
to modify the TP and SF for the end nodes’ uplink data
the users in the same channel. The simulation results indi-
transmissions. It should be noted that the network server
cated that the resource allocation mechanism obtained 80%
does not increase the SF (it does not reduce the data rate),
better performance than the baseline method, while being
as this is done by the LoRa node through the ADR-NODE
much less complex. The authors of [115] improved the basic
[100]. A major study on the introduction of new algorithms
ADR by dynamically designating the radio parameters, SF
for the allocation of SFs and TPs in LoRaWAN networks is
and TP, and using the OWA operator. This work focuses on
reported in [93], where two algorithms are presented, the first
increasing the network noise resilience and PDR in dense IoT
being the SensitivitySF allocation and the second being the
scenarios through recognition of the nature of OWA decision
AssignmentSF allocation, with the goal of maximizing the
making and PLR metric, reaching low energy consumption
throughput of each SF level and thus improving the overall
for all channel conditions.
network performance. The simulation results show that the
Fair Adaptive Data Rate (FADR) [117], The algorithm
proposed algorithms can significantly improve network per-
manages the allocation of SFs and TPs of the nodes, with the
formance when compared to the basic ADR strategy, with
goal of ensuring a fair data extraction rate between all nodes
the AssignmentSF algorithm , in particular, systematically
while limiting excessive TPs. It optimally combines SFs and
ensuring a high success rate for any traffic load incurred.
TP levels while also ensuring node longevity by limiting ex-
The authors of [111] propose a system for allocating SFs cessively high TP levels. Furthermore, simulations show that
and TPs in LoRaWAN networks. The goal of this system when applied to highly congested cells, the FADR is 300%
was to improve the packet error rate (PER) for users far more fair than the minimum airtime allocation method, while
from the base station and to make these networks more consuming nearly 22% less power. In [119], the authors ex-
equitable. The main idea behind this algorithm is to as- amined a novel resource allocation mechanism to adjust the
sign different SFs and control the power at different nodes allocation of SFs and TPs targeting the effects of co-SF and
to ensure that signals do not interfere with one another. inter-SF interference. The authors optimized the SF and TP
The PER for the overall network is reduced by 42% in allocations to maximize the average data rates. In addition,
simulation using this technique in NS-3, and the PER of the joint SF intractability and TP assignment problem was
end devices remote from the gateway is reduced by 50%. addressed by dividing them into two sub-problems: (i) SF
The authors of [112] defined a novel resource allocation assignment with fixed TP and (ii) TP assignment with fixed
mechanism to dynamically adjust the LoRaWAN CF and SF. From the simulation results, it is shown that the proposed
SF parameters for collisions reduction while increasing the mechanism improves the fairness, data rates, and throughput
PDR. Correspondingly, this work provided a heuristic to find performance when compared to the baseline algorithms. In
the optimal CF and SF parameters by investigating the RSSI [116], a resource allocation algorithm for LoRaWAN is in-
and the distance between the gateway and the IoT device. troduced, which involves energy harvesting. The authors also
By this means, the gateway could receive the transmitted provide a model to optimize SF allocation, energy harvesting
packet and with sufficient power in the selected SF value. duration and power devices for IoT transmission. The paper
Furthermore, it assigns more IoT devices to the lower SF presents two SF allocation algorithms using fairness or un-
values so as to reduce the interference. In terms of energy fairness of IoT devices. The simulation results showed that
consumption, the proposed scheme consumes 20% more than the unfair SF allocation algorithm maximized the minimum
the basic ADR. The authors of [113] presented a resource rate. Furthermore, the imperfect SF orthogonality did not af-
allocation scheme for raising the PDR by fine-tuning the fect the minimum performance rate. Finally, the authors came
LoRaWAN radio parameters. Specifically, a Mixed Integer to the conclusion that throughput performance is strongly
Linear Programming (MILP) formulation was implemented affected by co-SF interference interference and not by the
to achieve ideal values of SF and CF, taking into account energy deficiency.
the network traffic specifications. The authors explored the The authors in [118] focus on the LoRaWAN protocol
network traffic specifications to increase the DER along with since its modulation technology can use different SFs to
reducing the packet collision rate and energy consumption achieve more flexible communications, and they propose
in LoRaWAN. Another interesting study that introduced a a new resource allocation scheme based on Spatial Time
resource allocation mechanism as a joint TP and CF assign- Division Multiple Access that uses multi-layer virtual cells
ment optimization problem is presented in paper [114] called (STDMA), allowing for more efficient resource allocation.
low-complexity Matching Channel Assignment Algorithm A numerical analysis of the optimal power consumption to
(MCAA). It aims to ensure throughput fairness among IoT achieve the highest data rate is presented, allowing for the
devices, especially where multiple devices are connected. To adjustment of the cellular radius depending on the communi-
formulate channel assignment, they treated IoT devices and cation distance. The results show that the proposed model
channels as two sets of selfish players seeking to maximize outperforms the conventional system in terms of data rate
VOLUME 4, 2016 15
I. Cheikh et al.
Adjusted Allocation Energy

Parameter Scheme Efficiency
Improves the packet error rate for

users remote from the base station The average power Using high TP consumes more
SF and TP —— ⋆
[111] and to ensure that these networks consumption is high. energy.
are more fair.
Maximizes the throughput of each It is restricted by

The proposed system uses the
SensitivitySF spreading factor level hence sensitivity values and
SF and CR maximum TP, which results in ⋆
[93] allocation improving the overall performance associated service
high energy consumption.
of the network. radii.
It should be run at
regular intervals or
during certain events, The proposed system uses the
AssignmentSF Ensures a high success rate for any
SF and CR such as having a maximum TP, which results in ⋆
[93] allocation traffic load incurred.
number of EDs high energy consumption.
joining the
LoRaWAN cell.
An adaptive solution to manage the Consumes more energy

The proposed models
setting of the best LoRaWAN (almost 20% more), since they
CORRECT are more costly, since
parameters in order to reduce the allocate the devices
SF and CF and they allocate the ⋆
[112] channel utilization, packets proportionally, although it is
MARCO devices
collision and therefore maximize possible to allocate with a SF
proportionally.
the number of packets delivered. that consumes less energy.
General framework that considers

interference between users
transmitting at the same time on
The proposed system uses the
non-orthogonal waveform codes, User transmits with a
SF and TP —— maximum TP, which results in ⋆
[116] maximizes the minimum rate of maximum TP.
high energy consumption.
LoRa users harvesting energy from
an external source to allow
operation and uplink transmissions.
The scenario with the standard

Optimizes CF and SF pair No support for LoRaWAN assignment, which
assignment with an average 6.6% scenarios with longer was set to SF value 7, achieved
SF and CF MILP ⋆⋆
[113] increase in DER over the standard transmission distance 2.9 times lower energy
LoRaWAN assignment. and more gateways. consumption than the
proposed solution.
Energy consumption related to

power allocation is alleviated
A low-complexity channel in LoRa devices by
assignment algorithm to make the The proposed power implementing a gateway to
SF and CF MCAA LoRa network scalable by allocation scheme is a execute the power allocation ⋆⋆⋆
[114]
distributing the channel access centralized approach algorithm, achieving a good
decision making at the LoRa node. compromise between system
performance and battery life of
LoRa devices.
It has low power consumption

for all channel conditions.
In the proposed algorithm, Ordered However, for low noise
No evaluation of the
Weighted Averaging (OWA) is channels, ADR-OWA’s energy
effect of gateway
used as a decision making method consumption is slightly higher.
SF and TP ADR-OWA number on the ⋆⋆⋆
[115] able to consider the channel The reason is the use of a
performance of the
condition in the configuration of pessimistic approach, which
algorithm.
transmission parameters. configures the transmit power
of the nodes a little higher with
low noise.
Running at the network server,

With a highly loaded
using orthogonal different SFs and
network, the number Highly efficient in terms of
SF and TP ADR-NET different TP; Increase the global ⋆ ⋆ ⋆⋆
[100] of collisions is energy consumption.
network capacity; Extend the
increasing.
battery life of the nodes.
16 VOLUME 4, 2016
Manages the allocation of SFs and FADR is based on the

TPs of the nodes, ensures the assumption that all
Consuming nearly 22% less
SF and TP FADR longevity of the nodes by limiting nodes are initiated ⋆ ⋆ ⋆⋆
[117] energy.
excessively high transmission with the same
power levels. transmission power.
Energy efficiency
decreases as the
Multilayer
Offers efficient sharing of spectrum spreading factor
virtual cell Derives the optimal TP to
SF and TP to address the problem of reducing increases along with ⋆ ⋆ ⋆⋆
[118] based maximize energy efficiency.
the overall data rate of the network. the decrease in the
resource
number of virtual
cells.
The use of the

A joint SF and power allocation quadratic
scheme for maximization of the approximation does The energy savings offered by
minimum uplink throughput of the not necessarily the proposed power
SF and TP —— ⋆ ⋆ ⋆⋆
[119] end devices, subject to co-SF and guarantee a better optimization strategies reach
inter-SF interference, and power local optimum than up to 37.43%.
constraints. that offered by the
linear approximation.
APRA successfully improved

power consumption by 95%
An adaptive priority-based
No dynamic change and increased the battery
resource allocation mechanism to
SF, BW and of radio parameters in discharge time of the end
APRA enhance LoRaWAN scalability and ⋆ ⋆ ⋆⋆
[120] TP case of network device by up to 5 years, while
energy consumption in a dense IoT
modification. ensuring high packet delivery
scenario.
and low delay for high priority
applications.
TABLE 8. Energy efficiency of SFs allocation methods using multiple configurable parameters.
and power optimization. Paper [120] proposes An adaptive to improve the performance of MAC protocols in order to
priority-based resource allocation (APRA) mechanism to optimize them. This section investigates the energy efficiency
enhance LoRaWAN scalability and energy consumption in of the LoRaWAN MAC layer on both the network and node
a dense IoT scenario, simulations results shows that APRA sides.
successfully improved power consumption by 95% and in-
creased the battery discharge time of the end device by up to A. NETWORK SIDE
5 years, while ensuring high packet delivery and low delay 1) MAC protocols
for high priority applications. As previously stated, the MAC protocol used affects en-
Table 8 lists some of the SFs allocation strategies that ergy consumption. Lora MAC layer protocols are classi-
have been used to improve the performance of LoRaWAN fied into two groups: centralized-synchronous protocols and
networks and their energy consumption compared with the contention-based protocols. It has been demonstrated that
standard solution using multiple configurable parameters. the energy efficiency of synchronous protocols is 3 to 4
The authors of [121] propose a model for estimating the times greater than that of contention-based protocols [122].
life-time of LoRa monitoring nodes; they also evaluated the Below we present a collection of frequently cited works in
cost of battery replacement and damage penalty costs; for the literature that address the energy consumption of MAC
longer sensing intervals, the damage penalty is more signif- protocols :
icant than the cost of battery replacement. In addition to an In order to achieve a correct balance between energy
analysis of the use of energy from renewable energy sources consumption and network performance, the authors of [123]
available in the industrial environment, a cost-benefit study propose an Adaptive Duty Cycle Medium Access Control
of harvesting energy in terms of battery life and replacement (ADC-MAC) protocol. In fact, the LoRaWAN network uses
costs is included. the ALOHA protocol to allow access to the channel operating
without any restrictions imposed by the duty cycle, and there
IV. ENERGY EFFICIENCY AT MAC LAYER is no integration of node energy efficiency and traffic load.
The MAC layer is one of the most extensively investigated The proposed protocol’s main idea is to dynamically set the
research areas in the IoT network field. Based on our review node’s duty cycle by specifying three factors: node load, node
of the literature, several mechanisms have been proposed energy, and channel busy rate. The authors also make it clear
VOLUME 4, 2016 17
I. Cheikh et al.
that the wireless transceiver module consumes the most en- prove the scalability and operability of LoRa networks. In
ergy for the sensor node. In this regard, a wide range of duty order to evaluate the performance of the new MAC protocol,
cycle mechanisms in MAC protocols have been developed a comparison between DQ-LoRa, P-ALOHA, and CSMA
as an ideal solution for controlling the wireless transceiver was done, and the results showed remarkable performance
and, as a result, lowering energy consumption, some of these in terms of throughput, average delay, and average power
works are : synchronous duty cycle protocols (S-MAC [124], consumption. DQ-LoRa is more efficient than P-ALOHA in
T-MAC [125], EX-MAC [126]), asynchronous duty cycle terms of power consumption, and can achieve power savings
protocols (Wise-MAC [127], PW-MAC [128]). of up to 48% when the number of packets transmitted is high.
The authors of [40] propose an additional MAC protocol to In [133], the authors introduce, develop and evaluate FCA-
solve the problem of decoding superposed Lora signals in the LoRa, a new MAC protocol designed to improve reliability
case of orthogonal SFs with the same receiving power. Previ- and collision avoidance in LoRa networks. It is based on
ous research [129], [130], [61] has shown that signals are not the diffusion of beacon frames through the network gateway
completely orthogonal when using different SFs. While using aiming to synchronize communication with terminal devices
the same SF and channel, the stronger signal may be captured and to optimize channel utilization using different SFs. The
due to the difference in receiving power; otherwise, under simulation was performed using OMNeT++, showing that
similar receiving power conditions, a collision is generated FCA-LoRa increases the performance of the traditional Lo-
and all signals are considered lost. The proposed beacon- RaWAN scheduling method regarding throughput and colli-
based MAC protocol is the Collision Resolving-MAC (CR- sion avoidance. However, since the LoRa’s end devices listen
MAC) protocol, in which the collision resolution technique is almost continuously to the available frequency channels, the
used to decode the superposed Lora signals. Two algorithms FCA-LoRa could raise some serious issues regarding power
have been developed, the first for decoding two slightly consumption. Another study in which the authors proposed a
desynchronized superposed LoRa signals, and the second for real-time protocol called RT-LoRa for industrial monitoring
three or more, and the CRC field is used to improve the col- and control applications. The protocol uses a Multiple Listen-
lision resolution technique. The simulation results show that ing Before Talk (mLBT) mechanism that allows the detection
the CR-MAC protocol achieves significant improvements in of channels several times in a time slot [134].
terms of energy efficiency and latency levels. In [135], the authors present EF-LoRa, a LoRa network-
The authors of [131] propose a Carrier Sense Multiple ing solution that can provide equitable power consumption
Access (CSMA) protocol for a Lora network as an alternative between end devices based on smart allocation of various
to the random access protocol ALOHA, with the goal of network resources such as frequency channels, SFs, and
minimizing collisions in Lora transmissions for short and TP, ensuring balanced power consumption between LoRa
long messages, with extremely interesting results in terms of network end devices and extending the network’s life. Ac-
both power consumption and delay. cording to simulation results, the proposed EF-LoRa solution
A significant amount of effort was expended in [73] to can improve the energy fairness of existing LoRa networks
propose a new MAC protocol called RS-LoRa, with the goal by 177.8% and achieve higher energy efficiency fairness than
of improving the reliability and scalability of the LoRaWAN existing LoRa and RS-LoRa networks.
network. This RS-LoRa MAC protocol is divided into two The authors of [136] propose a new MAC protocol that
steps: the first is handled by the gateways, in which a gateway dynamically adapts the LoRaWAN MAC layer to changes in
schedules the nodes in its cell by dynamically specifying the traffic load. This protocol, known as the Traffic-aware En-
authorized received signal strength and SFs for each channel, ergy efficient MAC protocol (TREMA), can switch between
and the second is handled by the nodes, in which the nodes asynchronous and synchronous schemes based on changes
decide on the transmissions based on the scheduling infor- in probed traffic. TREMA eventually expands the maximum
mation provided by the gateway. The transmission power, SF, capacity of LoRa implementations while also ensuring that
channel, and timing of data transmission are all determined the most energy-efficient access scheme is always chosen.
by the nodes themselves. The nodes are divided into different A new MAC protocol, called Deterministic Group Ac-
groups using the proposed light scheduling, and each group knowledgment Transmissions in LoRa networks (DG-LoRa),
uses a common transmission power to reduce the capture designed to improve the scalability of loRaWAN networks
effect. This MAC protocol proposal significantly reduces the through deterministic GACK transmissions is introduced in
number of packet collisions in the network by improving [139]. The authors evaluate the performance of DG-LoRa
reliability and scalability. In terms of energy consumption, using a Monte Carlo simulation and then compare it to
it is clear that RS-LoRa introduces supplemental power con- existing LoRaWANs in terms of data drop rate and number
sumption measures. However, it should be noted that in RS- of retransmissions. Their numerical results show that DG-
LoRa, authorized SFs are determined by the gateway based LoRa supports about five times more connections to the LoRa
on network preferences; the gateway can choose between network by achieving a data drop rate of 5%. In addition,
reliability and energy consumption. The authors of [132] DG-LoRa provides low overhead by reducing the number
proposed including the distributed queueing (DQ) algorithm of data frame retransmissions. In [137], the authors propose
in the LoRa, called the DQ-based MAC protocol, to im- Time-slotted LoRaWAN (TS-LoRa) as a new approach to
18 VOLUME 4, 2016
Energy
Ref. MAC Protocols Advantages Drawbacks Energy Consumption
efficiency
In case of RS-LoRa
Improves the reliability and scalability
packet transmission, the
of the LoRaWAN network; reduces the
average power
number of packet collisions; SFs are
consumption is high. The The average power consumption is
RS-LoRa determined by the gateway according to ⋆
[73] gateway has the ability to high.
the network preferences, the gateway
choose between network
can make choices between reliability
reliability and power
and energy consumption.
consumption.
Listens most continuously

Synchronizes communication with
to the available frequency
terminal devices and optimizes channel The power consumption is high
channels, which could
FCA-LoRa utilization using different SFs. due to continuous listening to the ⋆
[133] lead to serious problems
increases the performance in terms of available frequency channels.
regarding power
throughput and collision avoidance.
consumption.
The device constantly

Uses a Multiple Listening Before Talk
checks the channels, Less efficient in terms of energy
(mLBT) mechanism that allows the
RT-LoRa which makes it less consumption due to continuous ⋆
[134] detection of channels several times in a
efficient in terms of channel checking.
time slot.
energy consumption.
Affected by packet
collisions and duty cycle
Gives the nodes sender the possibility
limitations, Pure-ALOHA P-ALOHA increases the battery
Pure-ALOHA to choose a SF, a channel and a ⋆⋆
[77] is more energy efficient life of the node when traffic is low.
transmission power.
when the traffic load is
low.
Unsuitable for
The channel time is divided into slots, Significantly reduces the collision
applications where high
Slotted-ALOHA end devices are only able to send rate, resulting in better energy ⋆⋆⋆
[78] reliability in data
packets at the beginning of a slot. efficiency.
transmission is needed.
Minimizes collisions in Lora

At high traffic loads, the proposed
transmissions for short and long Clearly increases sending
CSMA model has a significantly reduced ⋆⋆⋆
[131] messages; reduces power consumption latency.
energy consumption.
and delay.
Enhances the scalability and operability

Achieves power savings of up to
of LoRa networks; achieves power Less efficient in terms of
48% when the number of packets
DQ-based MAC savings of up to 48% when the number energy consumption than ⋆⋆⋆
[132] transmitted is high compared to the
of packets transmitted is high compared CSMA.
pure ALOHA protocol.
to the pure ALOHA protocol.
A higher number of nodes

indicated that TS-LoRa has lower
An efficient mechanism that allows The additional energy cost energy consumption than the
nodes to autonomously and securely of synchronization is the confirmable version of LoRaWAN
TS-LoRa deduce their slot number in the frame major issue when without incurring a compromise in ⋆⋆⋆
[137]
with high reliability and guaranteed designing synchronous packet delivery rate, whereas for
delay. communication protocols. confirmable traffic there is an
additional energy cost to maintain
acknowledgements.
In ADC-MAC protocol, each node

Achieves a correct balance between tries to avoid premature "death",
energy consumption and network thus extending the lifetime of the
performance ; dynamically set the duty Linked to the 3 indicators network. The main idea is to
ADC-MAC ⋆ ⋆ ⋆⋆
[123] cycle of the node by specifying 3 indicated. maintain the sleep state to extend
factors including : node load, the node lifetime of the node and to use
energy and the channel busy rate. residual energy as much as possible
to transmit emergency information.
Proposes a mathematical model of

energy efficiency in multi-gateway An increase in the
LoRa networks; Provides equitable convergence time of the
EF-LoRa can improve the energy
power consumption between end EF-LoRa algorithm linked
EF-LoRa efficiency of existing LoRa ⋆ ⋆ ⋆⋆
[135] devices based on a smart allocation of to the deployment of a
networks by 177.8%.
the various network resources, such as greater number of EDs
frequency channels, spreading factors and GWs.
and transmission power.
VOLUME 4, 2016 19
I. Cheikh et al.
The network reaches a round-trip

An energy-efficient network
latency of milliseconds, with the
architecture and a highly efficient The resulting system cost
sensing devices consuming less
On-demand on-demand TDMA communication increases slightly due to
than 46 mJ when active and 1.83 ⋆ ⋆ ⋆⋆
[138] TDMA protocol for IoT, addressing both the the addition of an extra
µW during inactive periods, and
energy efficiency and latency of wake-up receiver module.
lasting up to three years on a 1200
standard LoRa networks.
mAh lithium-polymer battery.
TREMA aims to dynamically

synchronize and desynchronize
LoRa deployments according to
Large simulation variations in traffic load, thus
Dynamic adaptation of the LoRaWAN
campaigns are needed to adapts network capacity and
MAC layer to traffic load variations,
establish the deployment always maximizes energy
TREMA increases the maximum achievable ⋆ ⋆ ⋆⋆
[136] fingerprint before efficiency. It offers good energy
throughput and maximizes the energy
implementing the efficiency when the generated
efficiency of the device.
mechanism. traffic is high. In contrast, pure
asynchronous ALOHA scheme is
more efficient when the generated
traffic is low.
Reduces the number of

The primary concern of
Supports about five times more retransmissions by more than three
DG-LoRa is to allocate
connections to the LoRa network by times compared to the legacy
SFs and time slots to
DG-LoRa achieving a 5% data drop rate, provides LoRaWAN network, enabling ⋆ ⋆ ⋆⋆
[139] maximize the number of
low overhead by reducing the number low-power operation of end
responses via GACK
of data frame retransmissions. devices and thus optimizing energy
transmission.
consumption.
The CR-MAC protocol

outperforms LoRaWAN with a
With a high number of
gain of up to 50% for small
Collision resolution technique is end devices, the risk of
networks and up to 90% for large
employed to decode the superposed multiple end devices
networks. Furthermore, using
CR-MAC Lora signals; CRC field is used to choosing the same sub-lot ⋆ ⋆ ⋆⋆
[40] SF12, the CR-MAC protocol
improve the collision resolution is high, thus reducing the
shows a large gain compared to
technique that has been adopted. CR-MAC coding rate and
LoRaWAN. The gain amounts to
overall throughput.
84% for small networks and 92%
for large networks.
TABLE 9. Energy efficiency of MAC protocols
time slot communication over LoRaWAN. TS-LoRa allows B. SENSOR/IOT DEVICE SIDE
nodes to self-organize the scheduling of time slots within The energy consumption of the end device is a very decisive
frames. Experimental results with 25 nodes show that TS- factor for the network’s correct and efficient performance
LoRa can achieve packet delivery rate above 99%, even [103], in the context of extremely rich and interesting work
for the most remote nodes. Furthermore, the simulations done to increase energy efficiency at the MAC layer level,
with a higher number of nodes indicated that TS-LoRa has including:
lower energy consumption than the confirmable version of
LoRaWAN, without compromising the packet delivery rate. 1) Scheduling of LoRa-based transmissions
The authors of [138] present a new network architecture and Authors of [66] propose a transmission scheduling algorithm
an on-demand Time Division Multiple Access (on-demand at a central node, which defines when a given IoT device
TDMA) MAC protocol exploiting short-range wake-up ra- is allowed to transmit. They reduce message size by imple-
dios and a LoRa physical layer. On-demand TDMA provides menting a probabilistic structure using Bloom filters, which
an efficient broadcast and unicast service for data transmis- encodes time slots designed to decrease the synchroniza-
sion and collection, thereby improving the performance of tion for packet length and send more information to IoT
LoRa networks and achieving 100% packet delivery rate by nodes. The time slots are assigned according to the traffic
eliminating the possibility of packet collisions. requirements of the IoT nodes and contextual information,
such as periodicity, synchronization or clock drift. Using the
Table 9 summarizes and evaluates the energy consumption central node, they synchronized the uplink transmission of
of different MAC protocols used in LoRaWAN networks. the IoT devices. In terms of power consumption, the results
show that the synchronization process will use less than
3 mAh of additional battery per end node over a period
20 VOLUME 4, 2016
of one year, for synchronization periods longer than three collection in LoRaWAN, and propose a FREE—fine-grained
days. This is less than the battery capacity used to transmit scheduling approach to ensure reliable and energy-efficient
packets that will be lost in an unsynchronized network due data collection in LoRaWAN networks. This approach is
to collisions. The authors of [140] present a scheme for based on buffering data at the terminal devices and col-
scheduling node communications in slots of different sizes lecting it in scheduled mass transmissions at appropriate
based on the SF. This approach enables transmissions with times. Instead of transmitting the data directly, this system
the same SF to be scheduled in other slots. On the other hand, assigns SFs, TPs, time slots, and frequency channels. Finally,
transmissions with different SFs can be processed in parallel, they evaluate the performance of the proposed system. The
thus preventing collisions. The first algorithm, called "Global numerical results show that the lifetime of the devices is
Algorithm", calculates the schedule for all communications. estimated to be more than ten years, regardless of the type
The second algorithm, called "Light", schedules only the of traffic and the size of the network.
first transmission for each individual node and replicates
it in subsequent frames. Under this setting, the IoT device 3) Data management
maintains the same SF in successive transmissions with the The frequency and payload size of data transmission affect
preference for the shortest schedule. The simulation results energy consumption; the authors of [145] add that for a given
show an improvement of up to 250% in terms of energy day, the number of transmissions is linked first and foremost
consumption associated with a packet delivery rate of nearly to the device’s energy consumption. As the payload size
100%. grows, so does the number of symbols per packet, resulting
In [141], a transmission scheduling mechanism based on in an additional energy cost per transmitted packet. It is also
temporal mappings is presented. The authors use the joining reported in [103] that the energy consumption in the case of
process of a new device to provide information about the unacknowledged uplink traffic is greater than when the ac-
periodization of transmissions. This enables gateways to knowledgement is transmitted in the first reception window.
schedule transmissions while avoiding collisions. The results In the primary use case, the end node without receiving ACK
obtained from NS-3 simulations show that, differently from must keep both reception windows open and the transition
LoRaWAN and CSMA, the collision rate decreases when the phase between the two, compared to the second case, where
packet delivery rate increases. with slightly increased energy the reception of the ACK packet is a one-time event. This
consumption. latter allows the node to enter sleep mode immediately after
receiving an ACK, and the transition phase between the first
2) Allocation scheme and second reception windows is avoided. Such a strategy
In [142], a new ADR algorithm called CA-ADR is proposed will aid in the reduction of energy consumption.
for the LoRaWAN network to assign data rates to EDs It is very clearly marked in [89] that each synchronization
by taking the collision probability at the MAC layer into procedure directly affects energy consumption through the
account. The new algorithm was compared to two benchmark excessive load of synchronization messages and their trans-
solutions using simulation and experimental approaches un- mission frequency, whereas LoRaWAN end devices are fre-
der different performance metrics. Their findings show that quently asked to share a common time base. This work con-
CA-ADR outperforms the standard solution in networks that sists of making a trade-off between the expected synchroniza-
are not severely constrained by connectivity issues. tion uncertainty and the energy available in the LoRaWAN.
Paper [143] proposes CARA (Collision Avoidance Re- On this basis, an algorithm was developed to generate com-
source Allocation), a new algorithm designed to increase parative curves between energy efficiency and uncertainty,
the capacity of LoRaWAN networks while decreasing the while taking into account a synchronization mechanism that
number of collisions. CARA divides the wireless medium’s occurs on demand following (a posteriori) synchronization.
capacity into resource blocks that correspond to a channel The trade-off proposal is applied to two scenarios related to
and a SF. Transmissions in different resource blocks will not industrial applications: the Time Division Multiple Access
collide due to the orthogonality of the SFs in LoRa. Further- (TDMA) system and the predictive maintenance framework.
more, CARA benefits from the existing joining procedure for The authors of [146] present Charm, a new system that
parameter exchange and synchronization, which eliminates aims to improve not only the battery life of end-devices but
any subsequent communication between the final devices and also the coverage of LPWAN networks. Charm is a solution
the network. At another point, a comparison was made be- designed to ensure that multiple LoRaWAN gateways pool
tween the ADR and the proposed SF assignment algorithm. their received signals in the cloud on a consistent basis for the
When compared to the ADR algorithm, CARA provides a purpose of detecting weak signals that can not be decoded by
significant increase in throughput. Another significant result any individual gateway. Indeed, a new gateway hardware and
is that, while the proposed solution slightly increases overall software design has been developed, consisting of precisely
transmission time, resulting in a slight increase in power con- detecting the specific sections of the received signal that must
sumption compared to the ADR proposed in the LoRaWAN be sent to the cloud. The obtained results provide intriguing
specification, it ensures more equitable resource sharing. benefits in terms of range and end node battery life.
The authors of [144] explore a different approach to data In [145], the authors propose an adaptive data aggregation
VOLUME 4, 2016 21
I. Cheikh et al.
Aggregation Energy
Technique Efficiency
Frame aggregation method for ACK,

LoRaWAN network server makes use of
ACKs that contain cumulative ACK
AggACK requires
information about multiple data and
[147] AggACK High level of latency. additional user nodes ⋆⋆
multiple users; It guarantees reliable data
power consumption.
transfer with a very small number of ACKs;
If a frame collision occurs, AggACK
ensures improved throughput.
With a data rate of 1, it is

Lead to a higher latency in
Adaptive data possible to guarantee an
Based on aggregating data from successive data transmission since
aggregation and extended lifetime of more
[145] periods and then transmitting the data using the data must be ⋆ ⋆ ⋆⋆
retransmission than five years for
a single LoRa packet. transmitted only after the
algorithm 6-minute time intervals
aggregation is complete.
with a 1000 mAh battery.
TABLE 10. Aggregation techniques used in LoRaWAN network.
and retransmission algorithm for transferring traffic data and thus data recovery from IoT applications is required.
from sensors, and the results were analyzed to provide an To accomplish this, they created and implemented the DaRe
estimated lifetime of sensor device radios. The proposed application-layer coding scheme. The need for such a system
algorithm works by aggregating data from successive periods is further motivated by the ALOHA media access technique
and transmitting it in a single LoRa packet. Because the data used in LoRaWAN, which will undoubtedly result in nu-
must be transmitted only after the aggregation is complete, merous collisions resulting in frame loss. DaRe is a system
such aggregation will result in higher latency in data trans- that incorporates new techniques to improve data recovery
mission. The reduction in energy costs associated with the while lowering overhead costs. DaRe does not recover lost
reduction in transmission frequency would result in increased frames, but it does allow data recovery from lost frames at
data latency. It has been demonstrated that even with a data the application level using FEC. The results show that with a
rate of one, a 1000 mAh battery can guarantee a lifetime of coding rate of 1/2 and a frame loss of up to 40%, 99% of the
more than five years at 6 minute time intervals. data is recoverable. In comparison to repeat encoding, DaRe
The authors of [147] present AggACK, a frame aggrega- provides 21% more data recovery and can reduce power
tion method for ACK running on LPWAN networks. The consumption during transmission by up to 42% for 10-byte
LoRaWAN network server employs the proposed method data units. Furthermore, DaRe provides greater resistance to
by sending ACKs that contain cumulative ACK information burst frame loss.
for multiple data and multiple users. User nodes open their
reception windows synchronously via the network server, 4) Unmanned Aerial Vehicles (UAVs)
causing the network server to simultaneously broadcast the Unmanned Aerial Vehicles (UAVs), also known as drones,
cumulative information of ACKs to the user nodes during are now being used in a wide range of novel applications,
the same ACK. Unlike the previous LoRaWAN, it ensures particularly in the telecommunications domain. Paper [149]
reliable data transfer with a very small number of ACKs. discusses some promising benefits of UAVs in wireless envi-
Table 10 lists the aggregation techniques used in Lo- ronments, such as UAV-assisted wireless charging, in which
RaWAN networks. a UAV could be (re)charged on the fly via wireless power
The LoRaWAN network in its basic configuration is un- transfer while flying near a charging station. A UAV, on
able to maintain reliable communication, and a loss of trans- the other hand, could wirelessly transfer energy to depleted
mitted frames is possible due to channel effects as well as ground IoT devices. At another level, authors of [150] present
terminal device mobility. The authors of [148] performed a novel use of UAV with an energy harvesting module to
a detailed measurement of a new LoRaWAN network to extend the network’s lifetime. As a result, the UAV can
identify the spatial and temporal properties of the LoRaWAN be used as an energy source for depleted IoT devices. On
channel. According to the collected data, frame losses are the one hand, the UAV charges the depletion ground IoT
very high and occur when end-devices move away from a devices, beginning with those with a battery level less than
gateway. When the end device is about 6 km from the nearest a certain threshold. The UAV station, on the other hand,
gateway, it can be as high as 53%. In fact, the loss of a frame collects data from IoT devices that have enough energy to
results in the loss of data. Given that an IoT application is transmit their packets, and during the same phase, the UAV
typically data-driven, the resulting data loss must be minimal, extracts and harvests energy from the RF signals transmitted
22 VOLUME 4, 2016
by IoT devices. the best choice between direct and multi-hop transmission is
Numerous studies, including [151], use UAVs on the Lo- determined by the sender-to-gateway distance. The authors
RaWAN network to collect data from LoRa sensors. The of [155] present an energy consumption model for single-
authors of [28] investigate the UAV-based gateway (GW) that hop and multi-hop LoRaWAN networks. In the multi-hop
can improve the reliability of LoRaWAN communication in approach, the authors consider a network formed by several
urban scenarios. However, neither study focuses on the power rings around the gateway. The simulations are performed
consumption and lifetime of the LoRaWAN terminal device. in MatLab, and the authors conclude that in the multi-hop
The authors of [152] highlight the energy efficiency of a scenario, nodes near the gateway consume more energy than
LoRaWAN using UAV technology to collect periodic sensor nodes farther away. In the single-hop scenario, however,
data reports. The simulation results show that a UAV-based nodes near the gateway have higher energy efficiency. One
GW can reduce the average power consumption for network more theoretical work has been introduced in [156] for var-
communications by up to 59%, depending on the trajectory ious multi-hop configurations for LoRa networks with three
or speed of the UAV. hops to the gateway. When compared to traditional single-
hop LoRaWAN, the results showed that some topologies
5) Linear wireless sensor network could improve packet delivery ratio and energy consumption.
The authors of paper [153] investigate the energy consump- The authors of [157] consider a standard LPWAN with a
tion of a specific type of WSN known as Linear WSN. TDMA MAC layer. They suggest a Distance-Ring Exponen-
According to [154], this type of WSN differs significantly tial Stations Generator (DRESG) framework for evaluating
from a standard WSN in that the sensor nodes are fixed performance and establishing optimal routing connections
in one dimension. Such WSNs are used for monitoring a in the uplink for multi-hop communications. Their results
wide range of applications, including transmission lines and indicate that multi-hop may improve network lifetime and
pipelines. Linear WSN requires a unique network archi- balance energy consumption across all nodes in the network.
tecture, and its power consumption varies. The authors of The same researchers suggest a protocol stack for LPWANs
[153] also proposed an energy consumption model based on called HARE in [158], which allows for single-hop and
a network architecture tailored for a linear wireless sensor multi-hop connections. It is formed of many techniques at
network, as well as a comparison of two widely used wireless various communication layers, such as network synchro-
communication protocols, LoRaWAN and ZigBee. Based on nization, adaptive transmission power level, TDMA chan-
this model, the LoRaWAN network outperforms the ZigBee nel access, network association process, and energy-aware
network in terms of energy efficiency. routing protocol. The protocol was tested on real hardware
platforms and demonstrated high reliability and low energy
V. ENERGY EFFICIENCY AT NET LAYER consumption.
The network layer’s primary function is to transmit data To improve the range and quality of LoRaWAN commu-
packets from a source to a given destination. This section nications, the authors of [159] propose the implementation
provides a literature review on LoRaWAN energy efficiency of a forwarding device that allows LoRaWAN to be used
at the network layer. Numerous authors proposed multi-hop for multi-hop communications. To assess the impact of an
LoRaWAN solutions where several devices act as relays for additional communication hop, tests were performed with
other devices. The routing mechanism is a crucial factor for Class A and Class C devices. It has been demonstrated that
a multi-hop LoRaWAN since it has the potential to affect the addition of a forwarder improves signal strength while
network performance in terms of throughput, reliability, la- significantly lowering the power consumption of the final
tency, and energy consumption. While some works propose device, resulting in a longer battery life. In [160], authors
mechanisms that use intermediate nodes, such as a simple propose using a simple relay device to expand the LoRaWAN
relay that only uses the LoRa physical layer, others propose coverage area in rural areas. The authors recommend deploy-
routing protocols, resulting in more complex mesh networks. ing relay nodes in areas that are not covered by the gateway
This section will go over these approaches in-depth. and propose a simple and direct message forwarder as well
as a synchronization mechanism. They demonstrate that the
A. NETWORK SIDE addition of the relay node to deliver packets reduces energy
1) Routing Models consumption. They also notice an improvement in coverage
As previously stated, the energy required to transmit a packet and reliability. The authors of [161] compare and contrast
decreases as the distance between devices decreases. The two architectures for multi-hop LoRaWAN in smart cities.
following is a collection of works that are more closely An end-device acts as a relay node in the first architecture
related to the routing model used for the LoRaWAN network: to extend the coverage area. In the second architecture,
The energy efficiency of the star and mesh topologies known as star-of-stars, a group of remote devices forms a
was investigated by the authors of [95]. They also propose cluster and sends data to a cluster gateway, which acts as a
a method for determining the best associations between relay and sends the data to a central gateway. The authors
spreading factor, transmission power, distance, and band- implement a prototype of both architectures, but no new
width. In terms of energy consumption, they conclude that forward mechanism is presented. The results show that two
VOLUME 4, 2016 23
I. Cheikh et al.
or three hops consume less energy than single-hop communi- and energy-efficient multi-hop communications. It is also
cation.Furthermore, this paper discusses some challenges as intended for low-latency and bidirectional communication.
well as future work for multi-hop LoRaWAN networks. It combines MAC and routing via beacons to enable time
In [84], a study on the implementation of a multi-hop synchronization and distance reporting based on the num-
network for LoRaWAN communication that provides low ber of hops to the gateway or sink. The authors of [162]
cost and high coverage is proposed. The proposed method propose using the Concurrent Transmission (CT) protocol in
involves sending packets to the Internet via multiple gate- LoRaWAN networks, resulting in the CT-LoRa protocol. CT
ways in order to overcome the problem of the gateway using is a new flood routing protocol that has been successfully
a private telecommunication network as a way to access implemented in IEEE 802.15.4 networks. The protocol does
the Internet, which requires a monthly fee. The network not require the use of a routing table, and the flooding
configuration includes : gateways, which are classified as mechanism ensures node synchronization. The proposed in-
hopping gateways (HGW), main gateways (MGW), and the door scenario results demonstrated that the protocol improves
network server. LoRa coverage and achieves a reliable packet delivery rate.
In terms of energy efficiency, we can conclude from the While the authors do not present an energy consumption
preceding works that the most energy efficient routing model analysis of the protocol, it is indicated that numerous works
is determined by the distance between the sender and the have confirmed that regardless of the energy consumption,
gateway. Table 11 provides a comparison between single-hop reliability, or latency viewpoint, CT provides better or com-
and multi-hop routing models. parable performance compared to state-of-the-art multi-hop
protocols. The authors of [85] provide a detailed analysis
Routing
Advantages Drawbacks of the definition and evaluation of a multi-hop routing pro-
model tocol based on the Destination-Sequenced Distance Vector
(DSDV) routing protocol, which is designed to increase the
Gateways require
Gateways located in the coverage of LoRaWAN installations and enable full interop-
simultaneous listening to
center communicate and
Single-hop transmit directly messages
all channels and erability with standard LoRaWAN gateways. The proposed
continuous connection to routing protocol makes use of the LoRaWAN beacon to
from the terminal devices
the Internet, which wastes
to the network server. deliver LoRaWAN packets from end nodes to gateways with
power.
the fewest hops possible. The proposed system was tested in
Improve the range and linear topologies and bottlenecks and evaluated using packet
quality of LoRaWAN Redundancies are
communications; expected in case of several receive rate (PRR) and throughput as performance metrics.
Multi-hop
Improves the signal transmissions reaching the Overall, the results indicate the viability of such an extended
strength and considerably gateway via different relay LoRaWAN multi-hop system. The throughput values ob-
reduces the power paths, resulting in a waste
consumption of the of battery power. tained are generally adequate for most IoT applications.
end-device. The authors of [165] describe a new version of the rout-
ing protocol for LoRa mesh networking, which provides
TABLE 11. Routing model for LoRaWAN network.
a multi-point networking connection between gateways to
achieve greater coverage. The proposed protocol is tailored
to the needs of LoRa networks and devices, and is based on
HWMP (Hybrid Wireless Mesh Protocol) and AODV (Ad
2) Routing Protocols Hoc Remote On-Demand Vector Routing). The protocol is
Routing is the process of looking for and selecting a path to evaluated only from the perspective of route construction
send messages from a source to a destination. In the presence time. The authors of [164] propose an energy-efficient multi-
of multiple routes, an algorithm manages path selection, hop communication solution (e2McH) in which routes are
taking into account one or more performance metrics to select built based on energy consumption, residual battery life, and
the links and nodes that will serve as the destination path. traffic rate. The simulation results demonstrate a 15% energy
Sensors are typically powered by small capacity batteries that consumption gain when compared to single-hop LoRaWAN.
operate autonomously over a limited lifetime, which can be In [166], the authors control the ancient underground water
extended using energy recovery techniques. As a result, one distribution systems in Siena, Italy, using multi-hop linear
of the primary challenges of WSN routing protocols is to communication over LoRa. The authors use a simple routing
create energy-efficient paths between sources and sinks [85]. protocol and synchronization mechanism, with end devices
Numerous studies, focusing on routing protocols, have using a wake-up time transmission scheme to save energy.
been conducted to improve the LoRaWAN network, as well They proved that the proposed solution is dependable, and
as: that the synchronization mechanism cuts energy consump-
In [38], an IoT protocol for LoRa transcievers called tion by half when compared to a non-optimal wake-up time.
LoRaBlink is proposed. This protocol addresses many issues In some applications, such as mines and pipelines, a linear
that are required for the deployment of IoT applications but topology may be the only viable option. In [163], the first
are not addressed by existing LoRa protocols, such as reliable proposal for a routing protocol in a LoRa network using a
24 VOLUME 4, 2016
Routing Energy
Ref. Working Concept Advantages Drawbacks Energy Consumption
Protocols Efficiency
Due to the obvious No power consumption

inevitable time and carrier analysis for the CT-LoRa
frequency offset between protocol is provided, it is
The protocol does not
The protocol transmitters, it is not stated that the CT protocol
require the use of a
improves LoRa applicable to all wireless consumes significantly
routing table, and the
[162] CT-LoRa coverage and standards. Even if the less power than other ⋆
flooding mechanism
achieves a reliable colliding packets have the multi-hop protocols due to
ensures node
packet delivery rate. same content, there are a the duty cycle accuracy
synchronization.
number of non-ideal enabled by the
effects that reduce well-programmed TDMA
receiver reliability. mechanism.
The proposed protocol is a
The proposed
simplified version of the
solution is feasible No energy consumption
DSDV routing protocol.
even for duty cycle The authors make the analysis, the authors
A simplified The proposed solution
limitations and the assumption that routing assume that routing nodes
[85] version of classifies nodes into two ⋆
achieved throughput nodes are not energy are powered while leaf
DSDV types: routing nodes and
values are satisfactory constrained. nodes are energy
leaf nodes, where the
for most IoT constrained.
routing devices are not
applications.
energy constrained.
It integrates MAC and
reliable and energy
routing, via beacons to If several transmissions To improve system power
efficient multi-hop
enable time reach the gateway via consumption, beacon
communications;
synchronization and different relaying routes, messages are sent
[38] LoRaBlink designed to support ⋆⋆
reporting the distance this results in redundancy infrequently. Although
low-latency and
based on the number of and therefore waste of this is not the most energy
bidirectional
hops to the gateway or battery power. efficient protocol.
communications.
sink.
Exploits control messages The proposed objective
that are distributed in the function used by the RPL
WSN, including the will be able to choose the
Link capacity limits do
sender’s rank on the root The protocol is not routing path that
not allow it to be
[85], node. which minimizes a based on advance minimizes ToA. By
RPL employed in a LoRaWAN ⋆⋆⋆
[163] certain measure such as knowledge of the choosing the path with the
setting in the absence of
power, number of hops, network topology. shortest ToA, power
significant adaptation.
number of expected consumption can be
transmissions (ETX) or reduced, extending the
other. network’s lifetime.
e2McH determines each
LoRa device’s energy
The proposed method
consumption, traffic rate, The simulation results
extends the battery
and residual battery life In the Line of Sight (LoS) demonstrate a 15% energy
life of LoRa devices
[164] e2McH and instructs a specific propagation link, all nodes consumption gain when ⋆ ⋆ ⋆⋆
in dense urban and
LoRa device to are evaluated. compared to single-hop
sub-urban
communicate in a LoRaWAN.
environments.
multi-hop configuration to
a nearby LoRa device.
TABLE 12. Energy efficiency of Routing Protocols.
standard IP stack was introduced. The developed a novel sensor nodes are associated with multi-hop communication,
MAC layer to deal with IPv6 Routing Protocol for Low the power consumption of the sensor nodes in hierarchical
Power and Lossy Network (RPL) routing protocol, and con- routing is extremely low [167].
cluded that enabling RPL over LoRa is a step in the right Table 12 examines a selection of protocols that are adapted
direction that needs to be better tested. It is demonstrated to the LoRaWAN network and are referenced in the literature.
that by selecting the path with the shortest ToA, power
consumption can be reduced, thereby increasing network 3) Network slicing
lifetime. For WSN, a multitude of routing protocols are Network slicing has emerged as one of the most important 5G
available. A hierarchical protocol is intended to achieve the innovations, gaining traction in both academia and industry
best possible balance of scalability and performance. When due to its ability to allow different IoT devices and appli-
VOLUME 4, 2016 25
I. Cheikh et al.
cations to coexist on top of a shared physical network infras- specifications [173]. The authors of [174] propose a novel
tructure. Network slicing, in general, makes use of virtualiza- distributed approach to solving the RAT selection problem,
tion and softwarization substrate technologies to dynamically based on matching game theory, which can achieve a stable
orchestrate physical networking resources (bandwidth, path- connected devices-RATs association. The proposed matching
ways, virtual network service chaining, placement). Network game assignment enables IoT devices with limited energy
slicing helps to create end-to-end virtual network instances budgets to improve their energy efficiency and reduce data
(networks slices) that are isolated on top of a shared physical transmission costs over the serving RAT while incurring very
network system [168], [169]. In this context, the literature in- little signaling overhead.
cludes several works for improving loRa energy consumption Around this perspective, we present below the most in-
through network slicing, such as: triguing works that make use of a Multi-RATs system with a
Authors of [170] study the factors associated with Iot LoRaWAN network and have demonstrated efficient energy
dense deployment scenarios, especially when faced with consumption :
the difficult task of using network slicing for a LoRaWAN Authors of [173] Explore the viability of developing a
gateway, as this results in performance degradation caused by multi-RAT LPWAN device capable of meeting transmis-
the physical limitations of these gateways. For the efficient sion requirements in an NB-IoT co-existence scenario. The
deployment and isolation of network slices in LoRaWAN authors built a device prototype that includes an NB-IoT
physical gateways, the authors used a software defined net- chipset that handles communication events and implements
working (SDN)-based architecture tailored to network slic- the LoRaWAN module. The dual-RAT LPWAN device was
ing. They also used a slice-based optimization method to tested in a commercial NB-IoT and a private LoRaWAN
improve the scalability and configuration of the LoRaWAN. environment at Brno University of Technology in the Czech
It employs TOPSIS and GMM algorithms to determine the Republic. According to the results, the NB-IoT chipset con-
best network slice configuration strategy for maximizing QoS sumed more than 200 mW on average to keep the network
benefits while minimizing energy consumption and reliabil- synchronized. The LoRaWAN transceiver, on the other hand,
ity costs. The numerical results demonstrate the efficiency of consumed less than 60 mW due to the lower transmission
the method used to enable optimal decision-making in real- power, despite the longer transmission time.
istic LoRaWAN scenarios for slice configuration than other The authors of [175] create an analytical model for as-
decision techniques (static, dynamic-adaptive, and dynamic- sessing the performance of grant-free IoT-based networks,
random). especially in scenarios where the same radio spectrum is used
The authors of [171] investigate network slicing over vari- by one or more concurrent radio access technologies (e.g.,
ous SF configurations in order to assess system performance LoRaWAN and Sigfox). This incited the authors performing
and determine which one best serves LoRa devices in each analytical modeling of transmission success probability and
slice. At a higher level, a dynamic inter-slicing technique is key performance indicators (KPIs) in terms of delay and bat-
designed in which the bandwidth is similarly reserved on all tery lifetime. It was possible to evaluate system performance
LoRa gateways based on a maximum likelihood estimation when subjected to interference by measuring the interference
(MLE) [172], and this is then improved and extended with of a scenario in which different communication technolo-
an adaptive dynamic method that evaluates each LoRa gate- gies coexist. The authors used simulations with analytical
way independently and reserves its bandwidth after applying models to evaluate the performance of a reference grant-free
MLE to the devices in its range. Both dynamic slicing sug- technology. Since exposed to interference caused by another
gestions will be compared to a simple fixed slicing strategy transmission device that uses the same frequency, the sce-
in which the GW’s bandwidth is reserved equally between nario was simulated using a simplified form of LoRaWAN.
slices. Finally, an energy model for LoRaWAN is integrated The simulation results show that the presence of compet-
into NS3 based on LoRa energy specifications to evaluate the ing technologies significantly degraded the system’s perfor-
energy consumed in each slice, and an intra-slicing algorithm mance. The authors solved this problem by demonstrating
that reaches the QoS requirements of each slice in isolation that it is possible to mitigate the interference caused by the
is proposed. The results demonstrate that as the number of coexistence of transmission devices through joint reception.
nodes increases, so does the total energy consumed across all The latter can significantly extend battery life, especially over
simulated slices. short distances.
In [176], the authors propose a game theory-based model
4) Multi-RATs integration in which end-nodes implement multiple radio transceivers
Gateways that allow multiple RATs (also known as multi- (e.g., LoRa, Wi-Fi, and BLE) to allow data processing at the
homing) have the potential to provide alternative commu- network edge and connect multiple networks using different
nication opportunities via various wireless networking tech- radio technologies. They integrate it into LoRaBox using
nologies, allowing IoT devices to send data. As a result, mul- auction-based techniques described in a previous work [177].
tihoming gateways must include control functions capable of The system is composed of a test solution that supports Lo-
selecting the best RAT interface for targeting IoT data based RaWAN, BLE, and Wi-Fi. Depending on the signal strength,
on a variety of factors such as current network conditions and radio, and application requirements, the system can switch
26 VOLUME 4, 2016
between multiple radio technologies adaptively. data and power.
B. SENSOR/IOT DEVICE SIDE VI. HOT TOPICS AND NEW INSIGHTS

Network Coding (NC) is a promising technique for increas- It is clear from the preceding sections that much-advanced
ing throughput and decreasing latency in the LoRaWAN work on improving energy efficiency within the LoRaWAN
network. Its main advantage is that it reduces the number network has already occurred. However, many issues remain
of transmissions, mostly the occurrence of packet loss in the unresolved to enable efficient energy use in the LoRaWAN
LoRaWAN network. The authors of [178] go into great detail network, leading us down an interesting research path. Table
about the methods and techniques used for Instantly Decod- 13 summarizes all axes addressed by layer and section, as
able Network Coding (IDNC), which is used in LoRaWAN. It well as the associated open research areas.
is possible to state that by simulating and implementing the
proposed system, the performance level in the case of joint A. DENSE/ULTRA-DENSE NETWORKS
delay, delivery time, and power consumption are reduced. Energy consumption control for the nodes is a primary issue
In paper [179] , two practical alternative methods for faced in a dense/ultra-dense network. This latter requires
increasing the probability of decoding a message in the a higher number of gateways and certainly a more dense
presence of interference are used. The first method employs deployment of end devices. Hence leading to more inter-
directional antennas to improve signal strength at the reference and consequently more collisions and delays. These
ceiver level at no additional cost to transmission power; the collisions have a significant impact on battery life.
second method employs multiple base stations to improve
the probability of decoding a message at least one receiver. B. SERVICE/DEVICE HETEROGENEITY
Both methods have advantages and disadvantages in terms Given the various types of devices used in a LoRaWAN net-
of cost, deployment method, and maintainability. The sim- work supporting a variety of services, managing and control-
ulation results show that, as expected, using each method ling the power consumption of heterogeneous devices from
has the potential to improve the performance of the LoRa different vendors would be necessary, posing a challenge
network under interference conditions. However, it has been to the dynamic integration of heterogeneous devices and
demonstrated that the use of multiple base stations outper- services, with systematic interoperability of multiple devices
forms the use of directional antennas. For example, in an in the network to be managed.
environment with 600 nodes interspersed through four other
LoRa networks of 600 nodes each, it is discovered that using C. CROSS-LAYER DESIGN
three base stations raises the Data Extraction Rate (DER) Cross-layer design is the process of designing protocols that
from 0.24 to 0.56, while using directional antennas raises the leverage the dependencies between protocol layers to achieve
value to 0.32. better performance. Contrary to layering, allowing the pro-
Some IoT applications, for example, may still require tocols of the different layers to be designed independently
certain reliability guarantees. The authors of paper [180] [183]. The idea in this context is to explore and create solid
approach this problem through a low complexity encod- and efficient cross-layer communication to improve network
ing/decoding technique based on fountain codes. The main performance, particularly those related to node energy con-
advantage of using encoding techniques when using LoRa sumption, by fully utilizing the interactions between the
technology is that no changes to the technology are required. layers.
The encoded message is transmitted by the end devices via
the radio access network to the application servers respon- D. MULTI-RATS ACCESS
sible for decoding the message. Each packet is assigned a The coexistence of heterogeneous technologies distributed
sequence number, which allows the application server to over various types of hardware is a research area that has
determine which packets have been lost. These packets can emerged due to the explosive growth of IoT devices. Mul-
then be recovered after a sufficient number of subsequent tiple radio access technologies (RATs) are highly future-
packets have been delivered successfully to ensure that the oriented wireless network technologies for improving coor-
established redundancy is used for decoding. Static Con- dination and collaboration between multi-RATs, while inte-
text Header Compression (SCHC) is a header compression grated multi-RAT architectures ensure that all RATs operate
scheme that allows for fragmentation and is particularly in a single virtual radio access network offering collabo-
suited to LPWAN technologies [181]. The authors of [182] rative interaction between RATs within the radio network.
discuss the impact of SCHC compression and fragmentation According to the reviewed literature, several researchers
on the LoRaWAN network, evaluating several compression have examined strategies to enable the coexistence of Lo-
and fragmentation configurations aimed at ensuring efficient RaWAN with different wireless technologies to satisfy the
packet transmission. The results demonstrate the benefits requirements of next-generation IoT applications that use
of SCHC, including a significant improvement in reliability LPWAN technologies. As a result, more significant gains
for LPWAN links operating at lower data rates. However, occur while preserving the benefits derived from indepen-
fragmentation can result in a loss of efficiency in terms of dent use of the network. By combining LoRaWAN with
VOLUME 4, 2016 27
I. Cheikh et al.
Section III : PHY Layer Section IV : MAC Layer Section V : NET Layer
• Configurable Radio Parameters • MAC Protocols. • Routing Models.

Selection. • Scheduling of LoRa-based • Routing Protocols.
• Ressource Allocation. transmissions. • Network slicing.
• Allocation scheme. • Multi-RATs interation.
• Data management.
• Unmanned Aerial Vehicles
(UAVs).
• Linear wireless sensor network.
Open Research Areas
• Improved energy consumption • Adaptive retransmission strategy. • Low power mesh network
through optimized settings. • Node energy management in solutions.
• A dynamic parameters selection dense/ultra-dense networks. • Energy / Performance trade-off.
systems to optimize energy • Machine learning and AI on the • Energy consumption and mobility.
consumption. edge.
• Cross technology interference.
TABLE 13. Summary of layered topics from the reviewed literature on LoRa and LoRaWAN and open research areas.
other wireless network technologies, multiple-RAT devices fect orthogonality. Hence the interest in developing scientific
can solve transmission bandwidth and rate-limiting issues. research that considers the non-orthogonality of SFs.
Multi-RAT can improve reliability by providing critical data Finally, the LoRaWAN network provides a high flexibility
transmission over multiple media at the same time. However, level: the NS can select the SF to be used by the various
the addition of multi-RAT devices holds the possibility of nodes, the transmission/reception channels, the duration of
introducing a significant new challenge: how do we deal with the reception windows. It is thus possible to support reliable
the interference caused by the coexistence of different tech- communications while meeting the requirements for com-
nologies that use the same wireless spectrum? Interference, munication reliability, delay, energy efficiency, and system
for example, causes information loss or packet retransmis- capacity with the proper configuration of these parameters.
sions, which affect various IoT applications, wastes energy, The task at hand is to implement a system that allows us
increase latency, and even reduce data rate. Several studies to automate the selection of LoRa transmission parameters
have highlighted the advantages of allowing LoRaWAN to based on the operating mode of each network to optimize
coexist with other LPWAN technologies, but they have failed and improve performance.
to address the issue of interference. Alternatively, further
research exploring the interference effects on large-scale and F. PACKETS RETRANSMISSIONS
heterogeneous LPWAN deployments is needed [13]. The ALOHA access scheme is used in the LoRaWAN net-
work, allowing any end device to send packets at any time by
simply making a link request. Given that terminal devices are
E. TRANSMISSION PARAMETERS SELECTION
deployed in a non-scheduled mode. And that the LoRaWAN
To begin with, the distribution of SF resources among nodes network uses a device-controlled communication mode, si-
is becoming a critical issue. Increasing the node’s SF should multaneous communications could potentially interfere with
be carefully calculated because higher SF means longer each other, resulting in packet transmission failure without a
ToA, and longer ToA increases the probability of collision suitable access scheme. Failure to communicate results in a
with other high-frequency transmissions. The challenge is decrease in energy efficiency. Investigating the probabilities
to propose a single SF assignment rule for each possible of retransmissions over the LoRaWAN network using various
LoRaWAN topology, given that each network is unique and SF allocation approaches based on the MAC protocol used to
requires different methods to optimize the SFs of its nodes. provide an analysis of the overall system performance is an
Furthermore, the studies conducted to solve this problem important direction to pursue.
use extremely high transmission power, resulting in higher
energy consumption. It is necessary to propose an allocation G. ENERGY / PERFORMANCE TRADE-OFF
scheme capable of optimizing SFs in an energy-efficient Most works targeting the performance improvement of Lo-
model. RaWAN networks focus on finding a good trade-off between
Secondly, the literature assumes that SFs are perfectly network performance and energy consumption. In several
orthogonal, with few studies examining the impact of imper- studies, the authors have to choose between a specific feature
28 VOLUME 4, 2016
to be improved (scalability, reliability, security, etc.) and [16] L. Beltramelli, A. Mahmood, P. Österberg, and M. Gidlund, “Lora
energy consumption. It is necessary to find a model that beyond aloha: An investigation of alternative random access protocols,”
IEEE Transactions on Industrial Informatics, vol. 17, no. 5, pp. 3544–
maintains a good equilibrium. 3554, 2020.
[17] R. Kufakunesu, G. P. Hancke, and A. M. Abu-Mahfouz, “A survey on
adaptive data rate optimization in lorawan: Recent solutions and major
VII. CONCLUSIONS challenges,” Sensors, vol. 20, no. 18, p. 5044, 2020.
This paper provides a comprehensive tutorial on LoRaWAN [18] J. R. Cotrim and J. H. Kleinschmidt, “Lorawan mesh networks: a review
networks. Namely, we provide a detailed background on the and classification of multihop communication,” Sensors, vol. 20, no. 15,
p. 4273, 2020.
Lora standard, its architecture, its communication protocol [19] L. Ntseane and B. Isong, “Analysis of lora/lorawan challenges,” in 2019
stack, its end devices, and its services. Then, we survey the International Multidisciplinary Information Technology and Engineering
main research works on energy efficiency at physical layer, Conference (IMITEC), pp. 1–7, IEEE, 2019.
[20] J. P. S. Sundaram, W. Du, and Z. Zhao, “A survey on lora networking:
MAC layer, and network layer of LoRa systems. Specifi- Research problems, current solutions, and open issues,” IEEE Communi-
cally, we highlight the most important work on the power cations Surveys & Tutorials, vol. 22, no. 1, pp. 371–388, 2019.
consumption of the LoRaWAN network, while focusing on [21] U. Raza, P. Kulkarni, and M. Sooriyabandara, “Low power wide area
networks: An overview,” IEEE Communications Surveys & Tutorials,
the network specific characteristics that are attributed to these vol. 19, no. 2, pp. 855–873, 2017.
specific layers. The goal of this tutorial is review pioneering [22] R. K. Singh, R. Berkvens, and M. Weyn, “Time synchronization with
research works and draw insights on how to optimize LoRa channel hopping scheme for lora networks,” in International Conference
on P2P, Parallel, Grid, Cloud and Internet Computing, pp. 786–797,
capabilities both from network perspective and device per- Springer, 2019.
spective. We finally present some research opportunities and [23] L. Li, J. Ren, and Q. Zhu, “On the application of lora lpwan technology in
open problems. sailing monitoring system,” in 2017 13th Annual Conference on Wireless
On-demand Network Systems and Services (WONS), pp. 77–80, IEEE,
2017.
REFERENCES [24] D. Hernandez, G. Peralta, L. Manero, R. Gomez, J. Bilbao, and C. Zubia,
“Energy and coverage study of lpwan schemes for industry 4.0,” in 2017
[1] G. A. Akpakwu, B. J. Silva, G. P. Hancke, and A. M. Abu-Mahfouz,
IEEE International Workshop of Electronics, Control, Measurement,
“A survey on 5g networks for the internet of things: Communication
Signals and their Application to Mechatronics (ECMSM), pp. 1–6, IEEE,
technologies and challenges,” IEEE Access, vol. 6, pp. 3619–3647, 2017.
2017.
[2] L. Chettri and R. Bera, “A comprehensive survey on internet of things
[25] S. Tomasin, S. Zulian, and L. Vangelista, “Security analysis of lorawan
(iot) towards 5g wireless systems,” IEEE Internet of Things Journal,
join procedure for internet of things networks,” in 2017 IEEE Wireless
2019.
Communications and Networking Conference Workshops (WCNCW),
[3] W. Ejaz, A. Anpalagan, M. A. Imran, M. Jo, M. Naeem, S. B. Qaisar, and pp. 1–6, IEEE, 2017.
W. Wang, “Internet of things (iot) in 5g wireless communications,” IEEE
[26] N. Naik, “Lpwan technologies for iot systems: choice between ultra
Access, vol. 4, pp. 10310–10314, 2016.
narrow band and spread spectrum,” in 2018 IEEE International Systems
[4] X. Liu, S. Wu, Y. Guo, and C. Chen, “The demand and development of in- Engineering Symposium (ISSE), pp. 1–8, IEEE, 2018.
ternet of things for 5g: A survey,” in 2018 IEEE International Conference
[27] W. Ayoub, A. E. Samhat, F. Nouvel, M. Mroue, and J.-C. Prévotet,
on Consumer Electronics-Taiwan (ICCE-TW), pp. 1–2, IEEE, 2018.
“Internet of mobile things: Overview of lorawan, dash7, and nb-iot
[5] O. Khutsoane, B. Isong, and A. M. Abu-Mahfouz, “Iot devices and appli- in lpwans standards and supported mobility,” IEEE Communications
cations based on lora/lorawan,” in IECON 2017-43rd Annual Conference Surveys & Tutorials, vol. 21, no. 2, pp. 1561–1581, 2018.
of the IEEE Industrial Electronics Society, pp. 6107–6112, IEEE, 2017. [28] V. Sharma, I. You, G. Pau, M. Collotta, J. Lim, and J. Kim, “Lorawan-
[6] J. Gubbi, R. Buyya, S. Marusic, and M. Palaniswami, “Internet of things based energy-efficient surveillance by drones for intelligent transporta-
(iot): A vision, architectural elements, and future directions,” Future tion systems,” Energies, vol. 11, no. 3, p. 573, 2018.
generation computer systems, vol. 29, no. 7, pp. 1645–1660, 2013. [29] K. Mekki, E. Bajic, F. Chaxel, and F. Meyer, “A comparative study of
[7] K. Kaur, “A survey on internet of things–architecture, applications, and lpwan technologies for large-scale iot deployment,” ICT express, vol. 5,
future trends,” in 2018 First International Conference on Secure Cyber no. 1, pp. 1–7, 2019.
Computing and Communication (ICSCCC), pp. 581–583, IEEE, 2018. [30] A. Lavric, A. I. Petrariu, and V. Popa, “Long range sigfox communication
[8] A. Lavric and V. Popa, “Internet of things and lora™ low-power wide- protocol scalability analysis under large-scale, high-density conditions,”
area networks: a survey,” in 2017 International Symposium on Signals, IEEE Access, vol. 7, pp. 35816–35825, 2019.
Circuits and Systems (ISSCS), pp. 1–5, IEEE, 2017. [31] M. Lombardi, F. Pascale, and D. Santaniello, “Internet of things: A
[9] Y. Song, J. Lin, M. Tang, and S. Dong, “An internet of energy things general overview between architectures, protocols and applications,”
based on wireless lpwan,” Engineering, vol. 3, no. 4, pp. 460–466, 2017. Information, vol. 12, no. 2, p. 87, 2021.
[10] M. H. M. Ghazali, K. Teoh, and W. Rahiman, “A systematic review of [32] H. Kwasme and S. Ekin, “Rssi-based localization using lorawan technol-
real-time deployments of uav-based lora communication network,” IEEE ogy,” IEEE Access, vol. 7, pp. 99856–99866, 2019.
Access, 2021. [33] M. T. Buyukakkaslar, M. A. Erturk, M. A. Aydin, and L. Vollero,
[11] P. Gkotsiopoulos, D. Zorbas, and C. Douligeris, “Performance deter- “Lorawan as an e-health communication technology,” in 2017 IEEE 41st
minants in lora networks: a literature review,” IEEE Communications Annual Computer Software and Applications Conference (COMPSAC),
Surveys & Tutorials, 2021. vol. 2, pp. 310–313, IEEE, 2017.
[12] S. Ugwuanyi, G. Paul, and J. Irvine, “Survey of iot for developing [34] K. A. Aldahdouh, K. A. Darabkh, W. Al-Sit, et al., “A survey of 5g
countries: performance analysis of lorawan and cellular nb-iot networks,” emerging wireless technologies featuring lorawan, sigfox, nb-iot and
Electronics, vol. 10, no. 18, p. 2224, 2021. lte-m,” in 2019 International Conference on Wireless Communications
[13] F. S. D. Silva, E. P. Neto, H. Oliveira, D. Rosário, E. Cerqueira, C. Both, Signal Processing and Networking (WiSPNET), pp. 561–566, IEEE,
S. Zeadally, and A. V. Neto, “A survey on long-range wide-area network 2019.
technology optimizations,” IEEE Access, 2021. [35] J. Petäjäjärvi, K. Mikhaylov, M. Pettissalo, J. Janhunen, and J. Iinatti,
[14] J. M. Marais, A. M. Abu-Mahfouz, and G. P. Hancke, “A survey on the “Performance of a low-power wide-area network based on lora technol-
viability of confirmed traffic in a lorawan,” IEEE Access, vol. 8, pp. 9296– ogy: Doppler robustness, scalability, and coverage,” International Jour-
9311, 2020. nal of Distributed Sensor Networks, vol. 13, no. 3, p. 1550147717699412,
[15] A. Raychowdhury and A. Pramanik, “Survey on lora technology: solution 2017.
for internet of things,” Intelligent Systems, Technologies and Applica- [36] K. Mikhaylov, J. Petäjäjärvi, and J. Janhunen, “On lorawan scalability:
tions, pp. 259–271, 2020. Empirical evaluation of susceptibility to inter-network interference,” in
VOLUME 4, 2016 29
I. Cheikh et al.
2017 European Conference on Networks and Communications (EuCNC), uation model for lora technology,” in 2015 14th International Conference
pp. 1–6, IEEE, 2017. on ITS Telecommunications (ITST), pp. 55–59, IEEE, 2015.
[37] H. Schroder Filho, J. Pissolato Filho, and V. Moreli, “The adequacy of [59] D. Croce, M. Gucciardo, S. Mangione, G. Santaromita, and I. Tinnirello,
lorawan on smart grids: A comparison with rf mesh technology,” in 2016 “Impact of lora imperfect orthogonality: Analysis of link-level perfor-
IEEE International Smart Cities Conference (ISC2), pp. 1–6, IEEE, 2016. mance,” IEEE Communications Letters, vol. 22, no. 4, pp. 796–799,
[38] M. Bor, J. E. Vidler, and U. Roedig, “Lora for the internet of things,” 2018.
2016. [60] A. Mahmood, E. Sisinni, L. Guntupalli, R. Rondón, S. A. Hassan, and
[39] J. Haxhibeqiri, A. Karaagac, F. Van den Abeele, W. Joseph, I. Moerman, M. Gidlund, “Scalability analysis of a lora network under imperfect
and J. Hoebeke, “Lora indoor coverage and performance in an industrial orthogonality,” IEEE Transactions on Industrial Informatics, vol. 15,
environment: Case study,” in 2017 22nd IEEE International Conference no. 3, pp. 1425–1436, 2018.
on Emerging Technologies and Factory Automation (ETFA), pp. 1–8, [61] A. Waret, M. Kaneko, A. Guitton, and N. El Rachkidy, “Lora throughput
IEEE, 2017. analysis with imperfect spreading factor orthogonality,” IEEE Wireless
[40] N. El Rachkidy, A. Guitton, and M. Kaneko, “Collision resolution Communications Letters, vol. 8, no. 2, pp. 408–411, 2018.
protocol for delay and energy efficient lora networks,” IEEE Transactions [62] L. Feltrin, C. Buratti, E. Vinciarelli, R. De Bonis, and R. Verdone, “Lo-
on Green Communications and Networking, vol. 3, no. 2, pp. 535–551, rawan: Evaluation of link-and system-level performance,” IEEE Internet
2019. of Things Journal, vol. 5, no. 3, pp. 2249–2258, 2018.
[41] J. de Carvalho Silva, J. J. Rodrigues, A. M. Alberti, P. Solic, and A. L. [63] L. Vangelista, L. Dell’Anna, and P. Palazzzoli, “A battery lifetime com-
Aquino, “Lorawan—a low power wan protocol for internet of things: A parison between lorawan and wireless mbus smart meters,” in 2019 IEEE
review and opportunities,” in 2017 2nd International Multidisciplinary International Black Sea Conference on Communications and Networking
Conference on Computer and Energy Science (SpliTech), pp. 1–6, IEEE, (BlackSeaCom), pp. 1–3, IEEE, 2019.
2017. [64] U. Noreen, A. Bounceur, and L. Clavier, “A study of lora low power
[42] S. Subashini, R. Venkateswari, and P. Mathiyalagan, “A study on lorawan and wide area network technology,” in 2017 International Conference on
for wireless sensor networks,” in Computing, Communication and Signal Advanced Technologies for Signal and Image Processing (ATSIP), pp. 1–
Processing, pp. 245–252, Springer, 2019. 6, IEEE, 2017.
[43] L. Casals, B. Mir, R. Vidal, and C. Gomez, “Modeling the energy [65] D. Bankov, E. Khorov, and A. Lyakhov, “On the limits of lorawan
performance of lorawan,” Sensors, vol. 17, no. 10, p. 2364, 2017. channel access,” in 2016 International Conference on Engineering and
[44] R. B. Sørensen, D. M. Kim, J. J. Nielsen, and P. Popovski, “Analysis of Telecommunication (EnT), pp. 10–14, IEEE, 2016.
latency and mac-layer performance for class a lorawan,” IEEE Wireless [66] J. Haxhibeqiri, I. Moerman, and J. Hoebeke, “Low overhead scheduling
Communications Letters, vol. 6, no. 5, pp. 566–569, 2017. of lora transmissions for improved scalability,” IEEE Internet of Things
[45] F. Adelantado, X. Vilajosana, P. Tuset-Peiro, B. Martinez, J. Melia- Journal, vol. 6, no. 2, pp. 3097–3109, 2018.
Segui, and T. Watteyne, “Understanding the limits of lorawan,” IEEE [67] E. Aras, G. S. Ramachandran, P. Lawrence, and D. Hughes, “Exploring
Communications magazine, vol. 55, no. 9, pp. 34–40, 2017. the security vulnerabilities of lora,” in 2017 3rd IEEE International
Conference on Cybernetics (CYBCONF), pp. 1–6, IEEE, 2017.
[46] A. Carlsson, I. Kuzminykh, R. Franksson, and A. Liljegren, “Measuring
[68] J. Kim and J. Song, “A dual key-based activation scheme for secure
a lora network: performance, possibilities and limitations,” in Internet
lorawan,” Wireless Communications and Mobile Computing, vol. 2017,
of Things, Smart Spaces, and Next Generation Networks and Systems,
2017.
pp. 116–128, Springer, 2018.
[69] S.-Y. Wang and T.-Y. Chen, “Increasing lorawan application-layer mes-
[47] M. Rizzi, P. Ferrari, A. Flammini, and E. Sisinni, “Evaluation of the
sage delivery success rates,” in 2018 IEEE Symposium on Computers and
iot lorawan solution for distributed measurement applications,” IEEE
Communications (ISCC), pp. 00148–00153, IEEE, 2018.
Transactions on Instrumentation and Measurement, vol. 66, no. 12,
[70] D. Bankov, E. Khorov, and A. Lyakhov, “Mathematical model of lorawan
pp. 3340–3349, 2017.
channel access with capture effect,” in 2017 IEEE 28th Annual Interna-
[48] J. Wenli, Z. Guoxin, B. Minqi, and J. Junchao, “Ultra-narrowband wire-
tional Symposium on Personal, Indoor, and Mobile Radio Communica-
less communication technology based on qvmck modulation,” in 2008
tions (PIMRC), pp. 1–5, IEEE, 2017.
China-Japan Joint Microwave Conference, pp. 185–187, IEEE, 2008.
[71] A.-I. Pop, U. Raza, P. Kulkarni, and M. Sooriyabandara, “Does bidirec-
[49] S. Hengstler, D. P. Kasilingam, and A. H. Costa, “A novel chirp modu-
tional traffic do more harm than good in lorawan based lpwa networks?,”
lation spread spectrum technique for multiple access,” in IEEE Seventh
in GLOBECOM 2017-2017 IEEE Global Communications Conference,
International Symposium on Spread Spectrum Techniques and Applica-
pp. 1–6, IEEE, 2017.
tions,, vol. 1, pp. 73–77, IEEE, 2002.
[72] F. Cuomo, M. Campo, A. Caponi, G. Bianchi, G. Rossini, and P. Pisani,
[50] F. Van den Abeele, J. Haxhibeqiri, I. Moerman, and J. Hoebeke, “Scala- “Explora: Extending the performance of lora by suitable spreading factor
bility analysis of large-scale lorawan networks in ns-3,” IEEE Internet of allocations,” in 2017 IEEE 13th International Conference on Wireless and
Things Journal, vol. 4, no. 6, pp. 2186–2198, 2017. Mobile Computing, Networking and Communications (WiMob), pp. 1–8,
[51] J. C. da Silva, D. d. L. Flor, V. A. de Sousa Junior, N. S. Bezerra, and A. A. IEEE, 2017.
de Medeiros, “A survey of lorawan simulation tools in ns-3,” Journal of [73] B. Reynders, Q. Wang, P. Tuset-Peiro, X. Vilajosana, and S. Pollin,
Communication and Information Systems, vol. 36, no. 1, pp. 17–30, 2021. “Improving reliability and scalability of lorawans through lightweight
[52] G. Boquet, P. Tuset-Peiro, F. Adelantado, T. Watteyne, and X. Vila- scheduling,” IEEE Internet of Things Journal, vol. 5, no. 3, pp. 1830–
josana, “Lr-fhss: Overview and performance analysis,” arXiv preprint 1842, 2018.
arXiv:2010.00491, 2020. [74] P. San Cheong, J. Bergs, C. Hawinkel, and J. Famaey, “Comparison of
[53] N. Sornin, M. Luis, T. Eirich, T. Kramp, and O. Hersent, “Rp2–1.0. 2 lorawan classes and their power consumption,” in 2017 IEEE Symposium
lorawan regional parameters,” LoRa Alliance TC and Others, 2020. on Communications and Vehicular Technology (SCVT), pp. 1–6, IEEE,
[54] N. C. Gaitan, “A long-distance communication architecture for medical 2017.
devices based on lorawan protocol,” Electronics, vol. 10, no. 8, p. 940, [75] L. Beltramelli, A. Mahmood, P. Österberg, M. Gidlund, P. Ferrari, and
2021. E. Sisinni, “Energy efficiency of slotted lorawan communication with
[55] T. T. Nguyen, H. H. Nguyen, R. Barton, and P. Grossetete, “Efficient out-of-band synchronization,” IEEE Transactions on Instrumentation
design of chirp spread spectrum modulation for low-power wide-area and Measurement, vol. 70, pp. 1–11, 2021.
networks,” IEEE Internet of Things Journal, vol. 6, no. 6, pp. 9503–9515, [76] L. Oliveira, J. J. Rodrigues, S. A. Kozlov, R. A. Rabêlo, and V. H. C. d.
2019. Albuquerque, “Mac layer protocols for internet of things: A survey,”
[56] N. Blenn and F. Kuipers, “Lorawan in the wild: Measurements from the Future Internet, vol. 11, no. 1, p. 16, 2019.
things network,” arXiv preprint arXiv:1706.03086, 2017. [77] T. Polonelli, D. Brunelli, and L. Benini, “Slotted aloha overlay on
[57] K. Mikhaylov, J. Petaejaejaervi, and T. Haenninen, “Analysis of capacity lorawan-a distributed synchronization approach,” in 2018 IEEE 16th In-
and scalability of the lora low power wide area network technology,” in ternational Conference on Embedded and Ubiquitous Computing (EUC),
European Wireless 2016; 22th European Wireless Conference, pp. 1–6, pp. 129–132, IEEE, 2018.
VDE, 2016. [78] T. Polonelli, D. Brunelli, A. Marzocchi, and L. Benini, “Slotted aloha
[58] J. Petajajarvi, K. Mikhaylov, A. Roivainen, T. Hanninen, and M. Pet- on lorawan-design, analysis, and deployment,” Sensors, vol. 19, no. 4,
tissalo, “On the coverage of lpwans: range evaluation and channel atten- p. 838, 2019.
30 VOLUME 4, 2016
[79] E. Altman, R. El Azouzi, and T. Jiménez, “Slotted aloha as a game with IEEE/IFIP Network Operations and Management Symposium, pp. 1–9,
partial information,” Computer networks, vol. 45, no. 6, pp. 701–713, IEEE, 2018.
2004. [101] T. Yatagan and S. Oktug, “Smart spreading factor assignment for lo-
[80] E.-A. Rachid, S. Essaid, J. Tania, B. El-Houssine, et al., “Modeling rawans,” in 2019 IEEE Symposium on Computers and Communications
slotted aloha as a stochastic game with random discrete power selection (ISCC), pp. 1–7, IEEE, 2019.
algorithms,” Journal of Computer Systems, Networks, and Communica- [102] R. Yasmin, J. Petäjäjärvi, K. Mikhaylov, and A. Pouttu, “On the in-
tions, vol. 2009, 2009. tegration of lorawan with the 5g test network,” in 2017 IEEE 28th
[81] I. Cheikh, E. Sabir, R. Aouami, M. Sadik, and S. Roy, “Throughput- Annual International Symposium on Personal, Indoor, and Mobile Radio
delay tradeoffs for slotted-aloha-based lorawan networks,” in 2021 Inter- Communications (PIMRC), pp. 1–6, IEEE, 2017.
national Wireless Communications and Mobile Computing (IWCMC), [103] J. Haxhibeqiri, E. De Poorter, I. Moerman, and J. Hoebeke, “A survey of
pp. 2020–2025, IEEE, 2021. lorawan for iot: From technology to application,” Sensors, vol. 18, no. 11,
[82] J.-T. Lim and Y. Han, “Spreading factor allocation for massive connec- p. 3995, 2018.
tivity in lora systems,” IEEE Communications Letters, vol. 22, no. 4, [104] H. Mroue, B. Parrein, S. Hamrioui, P. Bakowski, A. Nasser, E. M. Cruz,
pp. 800–803, 2018. and W. Vince, “Lora+: An extension of lorawan protocol to reduce
[83] J. Dias and A. Grilo, “Lorawan multi-hop uplink extension,” Procedia infrastructure costs by improving the quality of service,” Internet of
computer science, vol. 130, pp. 424–431, 2018. Things, p. 100176, 2020.
[84] M. Kim and J.-w. Jang, “A study on implementation of multi-hop net- [105] A. Tiurlikova, N. Stepanov, and K. Mikhaylov, “Method of assigning
work for lorawan communication,” in 2020 International Conference on spreading factor to improve the scalability of the lorawan wide area
Information Networking (ICOIN), pp. 553–555, IEEE, 2020. network,” in 2018 10th International Congress on Ultra Modern Telecom-
[85] J. Dias and A. Grilo, “Multi-hop lorawan uplink extension: specification munications and Control Systems and Workshops (ICUMT), pp. 1–4,
and prototype implementation,” Journal of Ambient Intelligence and IEEE, 2018.
Humanized Computing, vol. 11, no. 3, pp. 945–959, 2020. [106] A. Azari and C. Cavdar, “Self-organized low-power iot networks: A
[86] A. Gladisch, S. Rietschel, T. Mundt, J. Bauer, J. Goltz, and S. Wieden- distributed learning approach,” in 2018 IEEE Global Communications
mann, “Securely connecting iot devices with lorawan,” in 2018 Second Conference (GLOBECOM), pp. 1–7, IEEE, 2018.
World Conference on Smart Trends in Systems, Security and Sustainabil- [107] A. Peruzzo and L. Vangelista, “A power efficient adaptive data rate
ity (WorldS4), pp. 220–229, IEEE, 2018. algorithm for lorawan networks,” in 2018 21st International Symposium
[87] K.-H. Phung, H. Tran, Q. Nguyen, T. T. Huong, and T.-L. Nguyen, on Wireless Personal Multimedia Communications (WPMC), pp. 90–94,
“Analysis and assessment of lorawan,” in 2018 2nd International Con- IEEE, 2018.
ference on Recent Advances in Signal Processing, Telecommunications [108] A. Farhad, D.-H. Kim, and J.-Y. Pyun, “Resource allocation to massive
& Computing (SigTelCom), pp. 241–246, IEEE, 2018. internet of things in lorawans,” Sensors, vol. 20, no. 9, p. 2645, 2020.
[88] R. S. Sinha, Y. Wei, and S.-H. Hwang, “A survey on lpwa technology: [109] N. Benkahla, H. Tounsi, S. Ye-Qiong, and M. Frikha, “Enhanced adr
Lora and nb-iot,” Ict Express, vol. 3, no. 1, pp. 14–21, 2017. for lorawan networks with mobility,” in 2019 15th International Wireless
[89] M. Rizzi, A. Depari, P. Ferrari, A. Flammini, S. Rinaldi, and E. Sisinni, Communications & Mobile Computing Conference (IWCMC), pp. 1–6,
“Synchronization uncertainty versus power efficiency in lorawan net- IEEE, 2019.
works,” IEEE Transactions on Instrumentation and Measurement, [110] N. Benkahla, H. Tounsi, Y.-Q. Song, and M. Frikha, “Vhmm-based e-adr
vol. 68, no. 4, pp. 1101–1111, 2018. for lorawan networks with unknown mobility patterns,” in 2021 Inter-
[90] N. Harris, J. Curry, et al., “Development and range testing of a lorawan national Wireless Communications and Mobile Computing (IWCMC),
system in an urban environment,” International Journal of Electronics pp. 86–91, IEEE, 2021.
and Communication Engineering, vol. 12, no. 1, 2018. [111] B. Reynders, W. Meert, and S. Pollin, “Power and spreading factor
[91] R. Fujdiak, P. Mlynek, J. Misurec, and M. Strajt, “Simulated coverage es- control in low power wide area networks,” in 2017 IEEE International
timation of single gateway lorawan network,” in 2018 25th International Conference on Communications (ICC), pp. 1–6, IEEE, 2017.
Conference on Systems, Signals and Image Processing (IWSSIP), pp. 1– [112] J. Moraes, N. Matni, A. Riker, H. Oliveira, E. Cerqueira, C. Both, and
4, IEEE, 2018. D. Rosário, “An efficient heuristic lorawan adaptive resource allocation
[92] J. Ortín, M. Cesana, and A. Redondi, “Augmenting lorawan performance for iot applications,” in 2020 IEEE Symposium on Computers and
with listen before talk,” IEEE Transactions on Wireless Communications, Communications (ISCC), pp. 1–6, IEEE, 2020.
2019. [113] E. Sallum, N. Pereira, M. Alves, and M. Santos, “Improving quality-of-
[93] M. El-Aasser, T. Elshabrawy, and M. Ashour, “Joint spreading factor service in lora low-power wide-area networks through optimized radio
and coding rate assignment in lorawan networks,” in 2018 IEEE Global resource management,” Journal of Sensor and Actuator Networks, vol. 9,
Conference on Internet of Things (GCIoT), pp. 1–7, IEEE, 2018. no. 1, p. 10, 2020.
[94] M. Bor and U. Roedig, “Lora transmission parameter selection,” in [114] Z. Qin and J. A. McCann, “Resource efficiency in low-power wide-area
2017 13th International Conference on Distributed Computing in Sensor networks for iot applications,” in GLOBECOM 2017-2017 IEEE Global
Systems (DCOSS), pp. 27–34, IEEE, 2017. Communications Conference, pp. 1–7, IEEE, 2017.
[95] M. N. Ochoa, A. Guizar, M. Maman, and A. Duda, “Evaluating lora [115] J. Babaki, M. Rasti, and R. Aslani, “Dynamic spreading factor and power
energy efficiency for adaptive networks: From star to mesh topologies,” allocation of lora networks for dense iot deployments,” in 2020 IEEE 31st
in 2017 IEEE 13th International Conference on Wireless and Mobile Annual International Symposium on Personal, Indoor and Mobile Radio
Computing, Networking and Communications (WiMob), pp. 1–8, IEEE, Communications, pp. 1–6, IEEE, 2020.
2017. [116] F. Benkhelifa, Z. Qin, and J. McCann, “Minimum throughput maximiza-
[96] T. Bouguera, J.-F. Diouris, J.-J. Chaillout, R. Jaouadi, and G. Andrieux, tion in lora networks powered by ambient energy harvesting,” in ICC
“Energy consumption model for sensor nodes based on lora and lorawan,” 2019-2019 IEEE International Conference on Communications (ICC),
Sensors, vol. 18, no. 7, p. 2104, 2018. pp. 1–7, IEEE, 2019.
[97] N. Pukrongta and B. Kumkhet, “The relation of lorawan efficiency with [117] K. Q. Abdelfadeel, V. Cionca, and D. Pesch, “A fair adaptive data rate
energy consumption of sensor node,” in 2019 International Conference algorithm for lorawan,” arXiv preprint arXiv:1801.00522, 2018.
on Power, Energy and Innovations (ICPEI), pp. 90–93, IEEE, 2019. [118] Y. Kawamoto, R. Sasazawa, B. Mao, and N. Kato, “Multilayer virtual
[98] V. Talla, M. Hessar, B. Kellogg, A. Najafi, J. R. Smith, and S. Gollakota, cell-based resource allocation in low-power wide-area networks,” IEEE
“Lora backscatter: Enabling the vision of ubiquitous connectivity,” Pro- Internet of Things Journal, vol. 6, no. 6, pp. 10665–10674, 2019.
ceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous [119] L. Amichi, M. Kaneko, E. H. Fukuda, N. El Rachkidy, and A. Guitton,
Technologies, vol. 1, no. 3, pp. 1–24, 2017. “Joint allocation strategies of power and spreading factors with imperfect
[99] Y. Peng, L. Shangguan, Y. Hu, Y. Qian, X. Lin, X. Chen, D. Fang, and orthogonality in lora networks,” IEEE Transactions on Communications,
K. Jamieson, “Plora: A passive long-range data network from ambient vol. 68, no. 6, pp. 3750–3765, 2020.
lora transmissions,” in Proceedings of the 2018 Conference of the ACM [120] E. Lima, J. Moraes, H. Oliveira, E. Cerqueira, S. Zeadally, and
Special Interest Group on Data Communication, pp. 147–160, 2018. D. Rosário, “Adaptive priority-aware lorawan resource allocation for
[100] M. Slabicki, G. Premsankar, and M. Di Francesco, “Adaptive configura- internet of things applications,” Ad Hoc Networks, vol. 122, p. 102598,
tion of lora networks for dense iot deployments,” in NOMS 2018-2018 2021.
VOLUME 4, 2016 31
I. Cheikh et al.
[121] H. H. R. Sherazi, M. A. Imran, G. Boggia, and L. A. Grieco, “Energy [142] R. Marini, W. Cerroni, and C. Buratti, “A novel collision-aware adaptive
harvesting in lorawan: A cost analysis for the industry 4.0,” IEEE data rate algorithm for lorawan networks,” IEEE Internet of Things
Communications Letters, vol. 22, no. 11, pp. 2358–2361, 2018. Journal, 2020.
[122] G. Hassan, M. ElMaradny, M. A. Ibrahim, A. M. Rashwan, and H. S. [143] N. Chinchilla-Romero, J. Navarro-Ortiz, P. Muñoz, and P. Ameigeiras,
Hassanein, “Energy efficiency analysis of centralized-synchronous lora- “Collision avoidance resource allocation for lorawan,” Sensors, vol. 21,
based mac protocols,” in 2018 14th International Wireless Communica- no. 4, p. 1218, 2021.
tions & Mobile Computing Conference (IWCMC), pp. 999–1004, IEEE, [144] K. Q. Abdelfadeel, D. Zorbas, V. Cionca, and D. Pesch, “Free-fine-
2018. grained scheduling for reliable and energy efficient data collection in
[123] T. Deng, J. Zhu, and Z. Nie, “An improved lorawan protocol based on lorawan,” IEEE Internet of Things Journal, 2019.
adaptive duty cycle,” in 2017 IEEE 3rd Information Technology and [145] S. Mathur, A. Sankar, P. Prasan, and B. Iannucci, “Energy analysis
Mechatronics Engineering Conference (ITOEC), pp. 1122–1125, IEEE, of lorawan technology for traffic sensing applications,” in Intelligent
2017. Transportation Systems (ITS) World Congress, 2017.
[124] W. Ye, J. Heidemann, and D. Estrin, “An energy-efficient mac protocol [146] A. Dongare, R. Narayanan, A. Gadre, A. Luong, A. Balanuta, S. Kumar,
for wireless sensor networks,” in Proceedings. Twenty-First Annual Joint B. Iannucci, and A. Rowe, “Charm: exploiting geographical diversity
Conference of the IEEE Computer and Communications Societies, vol. 3, through coherent combining in low-power wide-area networks,” in 2018
pp. 1567–1576, IEEE, 2002. 17th ACM/IEEE International Conference on Information Processing in
[125] T. Van Dam and K. Langendoen, “An adaptive energy-efficient mac Sensor Networks (IPSN), pp. 60–71, IEEE, 2018.
protocol for wireless sensor networks,” in Proceedings of the 1st interna- [147] Y. Hasegawa and K. Suzuki, “A multi-user ack-aggregation method for
tional conference on Embedded networked sensor systems, pp. 171–180, large-scale reliable lorawan service,” in ICC 2019-2019 IEEE Interna-
ACM, 2003. tional Conference on Communications (ICC), pp. 1–7, IEEE, 2019.
[126] Y.-X. Li, H.-S. Shi, and S.-P. Zhang, “An energy-efficient mac protocol [148] P. J. Marcelis, V. S. Rao, and R. V. Prasad, “Dare: Data recovery through
for wireless sensor network,” in 2010 3rd International Conference on application layer coding for lorawan,” in 2017 IEEE/ACM Second Inter-
Advanced Computer Theory and Engineering (ICACTE), vol. 4, pp. V4– national Conference on Internet-of-Things Design and Implementation
619, IEEE, 2010. (IoTDI), pp. 97–108, IEEE, 2017.
[127] A. El-Hoiydi and J.-D. Decotignie, “Wisemac: An ultra low power [149] L. Abouzaid, E. Sabir, H. Elbiaze, A. Errami, and O. Benhmammouch,
mac protocol for multi-hop wireless sensor networks,” in International “The meshing of the sky: Delivering ubiquitous connectivity to ground
symposium on algorithms and experiments for sensor systems, wireless internet of things,” IEEE Internet of Things Journal, vol. 8, no. 5,
networks and distributed robotics, pp. 18–31, Springer, 2004. pp. 3743–3757, 2020.
[128] L. Tang, Y. Sun, O. Gurewitz, and D. B. Johnson, “Pw-mac: An energy- [150] S. Arabi, H. Elbiaze, E. Sabir, and M. Sadik, “Tradeoffs for data col-
efficient predictive-wakeup mac protocol for wireless sensor networks,” lection and wireless energy transfer dilemma in iot environments,” in
in 2011 Proceedings IEEE INFOCOM, pp. 1305–1313, IEEE, 2011. 2018 IEEE International Conference on Communications (ICC), pp. 1–6,
IEEE, 2018.
[129] D. Croce, M. Gucciardo, I. Tinnirello, D. Garlisi, and S. Mangione,
“Impact of spreading factor imperfect orthogonality in lora communica- [151] D. Carrillo and J. Seki, “Rural area deployment of internet of things
tions,” in International Tyrrhenian Workshop on Digital Communication, connectivity: Lte and lorawan case study,” in 2017 IEEE XXIV Interna-
pp. 165–179, Springer, 2017. tional Conference on Electronics, Electrical Engineering and Computing
(INTERCON), pp. 1–4, IEEE, 2017.
[130] G. Zhu, C.-H. Liao, M. Suzuki, Y. Narusue, and H. Morikawa, “Eval-
uation of lora receiver performance under co-technology interference,” [152] A. Tiurlikova, N. Stepanov, and K. Mikhaylov, “Improving the energy ef-
in 2018 15th IEEE Annual Consumer Communications & Networking ficiency of a lorawan by a uav-based gateway,” in 2019 11th International
Conference (CCNC), pp. 1–7, IEEE, 2018. Congress on Ultra Modern Telecommunications and Control Systems and
Workshops (ICUMT), pp. 1–6, IEEE, 2019.
[131] C. Pham, “Investigating and experimenting csma channel access mech-
[153] A. Kurtoglu, J. Carletta, and K.-S. Lee, “Energy consumption in long-
anisms for lora iot networks,” in 2018 IEEE Wireless Communications
range linear wireless sensor networks using lorawan and zigbee,” in 2017
and Networking Conference (WCNC), pp. 1–6, IEEE, 2018.
IEEE 60th International Midwest Symposium on Circuits and Systems
[132] W. Wu, Y. Li, Y. Zhang, B. Wang, and W. Wang, “Distributed queueing (MWSCAS), pp. 1163–1167, IEEE, 2017.
based random access protocol for lora networks,” IEEE Internet of Things
[154] C. W. Chen and Y. Wang, “Chain-type wireless sensor network for
Journal, 2019.
monitoring long range infrastructures: architecture and protocols,” Inter-
[133] A. Triantafyllou, P. Sarigiannidis, T. Lagkas, I. D. Moscholios, and A. Sa- national Journal of Distributed Sensor Networks, vol. 4, no. 4, pp. 287–
rigiannidis, “Leveraging fairness in lorawan: A novel scheduling scheme 314, 2008.
for collision avoidance,” Computer Networks, vol. 186, p. 107735, 2021.
[155] M. S. Iqbal, G. W. Wiriasto, et al., “Multi-hop uplink for low power wide
[134] Q. L. Hoang, W.-S. Jung, T. Yoon, D. Yoo, and H. Oh, “A real-time lora area networks using lora technology,” in 2019 6th International Confer-
protocol for industrial monitoring and control systems,” IEEE Access, ence on Information Technology, Computer and Electrical Engineering
vol. 8, pp. 44727–44738, 2020. (ICITACEE), pp. 1–5, IEEE, 2019.
[135] W. Gao, W. Du, Z. Zhao, G. Min, and M. Singhal, “Towards energy- [156] M. O. Farooq, “Introducing scalability in lora-based networks through
fairness in lora networks,” in 2019 IEEE 39th International Conference multi-hop communication setups,” in 2019 IEEE Global Communica-
on Distributed Computing Systems (ICDCS), pp. 788–798, IEEE, 2019. tions Conference (GLOBECOM), pp. 1–6, IEEE, 2019.
[136] L. Chasserat, N. Accettura, B. Prabhu, and P. Berthou, “Trema: A traffic- [157] S. Barrachina-Muñoz, B. Bellalta, T. Adame, and A. Bel, “Multi-hop
aware energy efficient mac protocol to adapt the lorawan capacity,” 2021. communication in the uplink for lpwans,” Computer Networks, vol. 123,
[137] D. Zorbas, K. Abdelfadeel, P. Kotzanikolaou, and D. Pesch, “Ts-lora: pp. 153–168, 2017.
Time-slotted lorawan for the industrial internet of things,” Computer [158] T. Adame Vázquez, S. Barrachina-Muñoz, B. Bellalta, and A. Bel, “Hare:
Communications, vol. 153, pp. 1–10, 2020. Supporting efficient uplink multi-hop communications in self-organizing
[138] R. Piyare, A. L. Murphy, M. Magno, and L. Benini, “On-demand lora: lpwans,” Sensors, vol. 18, no. 1, p. 115, 2018.
Asynchronous tdma for energy efficient and low latency communication [159] B. Van de Velde, “Multi-hop lorawan: including a forwarding node,”
in iot,” Sensors, vol. 18, no. 11, p. 3718, 2018. [160] D. Mamour and P. Congduc, “Increased flexibility in long-range iot
[139] J. Lee, Y. S. Yoon, H. W. Oh, and K. R. Park, “Dg-lora: Deterministic deployments with transparent and light-weight 2-hop lora approach,” in
group acknowledgment transmissions in lora networks for industrial iot 2019 Wireless Days (WD), pp. 1–6, IEEE, 2019.
applications,” Sensors, vol. 21, no. 4, p. 1444, 2021. [161] M. S. Aslam, A. Khan, A. Atif, S. A. Hassan, A. Mahmood, H. K.
[140] D. Zorbas, K. Q. Abdelfadeel, V. Cionca, D. Pesch, and B. O’Flynn, Qureshi, and M. Gidlund, “Exploring multi-hop lora for green smart
“Offline scheduling algorithms for time-slotted lora-based bulk data cities,” IEEE Network, vol. 34, no. 2, pp. 225–231, 2019.
transmission,” in 2019 IEEE 5th World Forum on Internet of Things (WF- [162] C.-H. Liao, G. Zhu, D. Kuwabara, M. Suzuki, and H. Morikawa, “Multi-
IoT), pp. 949–954, IEEE, 2019. hop lora networks enabled by concurrent transmission,” IEEE Access,
[141] T.-H. To and A. Duda, “Timemaps for improving performance of lo- vol. 5, pp. 21430–21446, 2017.
rawan,” in ICC 2020-2020 IEEE International Conference on Commu- [163] B. Sartori, S. Thielemans, M. Bezunartea, A. Braeken, and K. Steenhaut,
nications (ICC), pp. 1–7, IEEE, 2020. “Enabling rpl multihop communications based on lora,” in 2017 IEEE
32 VOLUME 4, 2016
13th International Conference on Wireless and Mobile Computing, Net- IMANE CHEIKH (S’19) received the B.Sc. de-
working and Communications (WiMob), pp. 1–8, IEEE, 2017. gree in Networks and telecommunications from
[164] M. Anedda, C. Desogus, M. Murroni, D. D. Giusto, and G.-M. Muntean, Faculty of Sciences and Technics (FSTS), Hassan
“An energy-efficient solution for multi-hop communications in low power first University, Settat, Morocco, in 2016, and the
wide area networks,” in 2018 IEEE International Symposium on Broad- M.Sc. degree in computer science and logistics
band Multimedia Systems and Broadcasting (BMSB), pp. 1–5, IEEE, from Higher School of Textile and Clothing Indus-
2018. tries (ESITH), Casablanca, Morocco, in 2018. She
[165] D. Lundell, A. Hedberg, C. Nyberg, and E. Fitzgerald, “A routing
is currently pursuing the Ph.D. degree at the Na-
protocol for lora mesh networks,” in 2018 IEEE 19th International
tional Higher School of Electricity and Mechanics
Symposium on" A World of Wireless, Mobile and Multimedia Net-
works"(WoWMoM), pp. 14–19, IEEE, 2018. (ENSEM), Hassan II University of Casablanca,
[166] A. Abrardo and A. Pozzebon, “A multi-hop lora linear sensor network for Networking, Embedded Systems and Telecom Research Group (NEST). Her
the monitoring of underground environments: The case of the medieval current research interests include wireless networking and mobile commu-
aqueducts in siena, italy,” Sensors, vol. 19, no. 2, p. 402, 2019. nications (5G and beyond), the IoT, performance evaluation, smart cities,
[167] A. M. Zungeru, L.-M. Ang, and K. P. Seng, “Classical and swarm energy efficiency and low-energy networks.
intelligence based routing protocols for wireless sensor networks: A
survey and comparison,” Journal of Network and Computer Applications,
vol. 35, no. 5, pp. 1508–1536, 2012.
[168] S. E. Elayoubi, S. B. Jemaa, Z. Altman, and A. Galindo-Serrano, “5g
ran slicing for verticals: Enablers and challenges,” IEEE Communications
Magazine, vol. 57, no. 1, pp. 28–34, 2019.
RACHID AOUAMI (S’12, M’16) received his
[169] I. Afolabi, T. Taleb, K. Samdanis, A. Ksentini, and H. Flinck, “Network
slicing and softwarization: A survey on principles, enabling technologies, M.S and Ph. D. degrees in electrical engineering
and solutions,” IEEE Communications Surveys & Tutorials, vol. 20, no. 3, from ENSEM, Hassan 2 University, Casablanca,
pp. 2429–2453, 2018. Morocco, in 2012 and 2017 respectively. Since
[170] S. Dawaliby, A. Bradai, and Y. Pousset, “Network slicing optimization 2018, he has been a Postdoctoral fellow in Electri-
in large scale lora wide area networks,” in 2019 IEEE Conference on cal Computer Science at 3IT, Université de Sher-
Network Softwarization (NetSoft), pp. 72–77, IEEE, 2019. brooke, Quebec, Canada. Most recently, Rachid
[171] S. Dawaliby, A. Bradai, and Y. Pousset, “Adaptive dynamic network has been a research assistant with the RioTinTo
slicing in lora networks,” Future generation computer systems, vol. 98, and Sherbrooke University on projects from im-
pp. 697–707, 2019. proving Health care diagnoses to making au-
[172] S. Dawaliby, A. Bradai, Y. Pousset, and R. Riggio, “Dynamic network tonomous Wireless devices and robots. He is current research activities are in
slicing for lorawan,” in 2018 14th International Conference on Network performance analysis, power saving algorithms, and optimization in wireless
and Service Management (CNSM), pp. 134–142, IEEE, 2018. sensor networks. He is a member of the IEEE.
[173] K. Mikhaylov, M. Stusek, P. Masek, V. Petrov, J. Petajajarvi, S. Andreev,
J. Pokorny, J. Hosek, A. Pouttu, and Y. Koucheryavy, “Multi-rat lpwan
in smart cities: Trial of lorawan and nb-iot integration,” in 2018 IEEE
International Conference on Communications (ICC), pp. 1–6, IEEE,
2018.
[174] S. Arabi, H. El Hammouti, E. Sabir, H. Elbiaze, and M. Sadik, “Rat
association for autonomic iot systems,” IEEE Network, vol. 33, no. 6, ESSAID SABIR (S’07, M’10, SM’14) received
pp. 116–123, 2019. the B.Sc. degree in electrical engineering and au-
[175] M. Masoudi, A. Azari, E. A. Yavuz, and C. Cavdar, “Grant-free radio tomation from Mohammed V University of Rabat,
access iot networks: Scalability analysis in coexistence scenarios,” in Morocco, in 2004, the M.Sc. degree in telecommu-
2018 IEEE International Conference on Communications (ICC), pp. 1–7, nications and wireless engineering from the Na-
IEEE, 2018. tional Institute of Posts and Telecommunications,
[176] M. Haghighi, Z. Qin, D. Carboni, U. Adeel, F. Shi, and J. A. McCann, in 2007, the Ph.D. degree (Hons.) in networking
“Game theoretic and auction-based algorithms towards opportunistic
and computer engineering from University of Avi-
communications in lpwa lora networks,” in 2016 IEEE 3rd World Forum
gnon, France, in 2010, and the Habilitation degree
on Internet of Things (WF-IoT), pp. 735–740, IEEE, 2016.
[177] K. Maraslis, T. Spyridopoulos, G. Oikonomou, T. Tryfonas, and (Hons.) from Hassan II University of Casablanca,
M. Haghighi, “Application of a game theoretic approach in smart sensor in 2014. He was a contractual Assistant Professor at the University of Avi-
data trustworthiness problems,” in IFIP International Information Secu- gnon, from 2009 to 2012. Currently, he is a Full Professor at the department
rity and Privacy Conference, pp. 601–615, Springer, 2015. of electrical engineering at the National Higher School of Electricity and
[178] S. M. Joans and G. Aishwarya, “Joint delay and energy minimization Mechanics (ENSEM) where he leads the Networking Embedded Systems
using instantly decodable network coding for lorawan,” 2018. and Telecom research Group. He is/was the main investigator of many
[179] T. Voigt, M. Bor, U. Roedig, and J. Alonso, “Mitigating inter-network national projects, and has been involved in several other national and inter-
interference in lora networks,” arXiv preprint arXiv:1611.00688, 2016. national/European projects. His research interests include wireless networks
[180] M. Sandell and U. Raza, “Application layer coding for iot: benefits, and mobile communications, 5G and beyond, the IoT, infrastructure-less
limitations, and implementation aspects,” IEEE Systems Journal, vol. 13, networking, ultra-reliable low latency communications, wireless artificial
no. 1, pp. 554–561, 2018. intelligence and networking games. His work has been awarded in many
[181] W. Ayoub, M. Mroue, F. Nouvel, A. E. Samhat, and J.-C. Prévotet, international venues (IEEE 5GWF’21, IEEE IWCMC’19, IEEE WF-IoT’20,
“Towards ip over lpwans technologies: Lorawan, dash7, nb-iot,” in 2018 etc.). His Ph.D. thesis has been funded by an excellence scholarship from
sixth international conference on digital information, networking, and
INRIA-France. He was also a recipient of the excellence scholarship from
wireless communications (dinwc), pp. 43–47, IEEE, 2018.
the Moroccan Centre for Scientific and Technical Research. As an attempt to
[182] J. Sanchez-Gomez, J. Gallego-Madrid, R. Sanchez-Iborra, J. Santa, and
A. F. Skarmeta, “Impact of schc compression and fragmentation in lpwan: bridge the gap between academia and industry, he founded the International
A case study with lorawan,” Sensors, vol. 20, no. 1, p. 280, 2020. Conference on Ubiquitous Networking (UNet) and co-founded the Inter-
[183] V. Srivastava and M. Motani, “Cross-layer design: a survey and the road national Conference on Wireless Networks and Mobile Communications
ahead,” IEEE communications magazine, vol. 43, no. 12, pp. 112–119, (WINCOM). He serves as an Editor, Associate Editor, and/or Reviewer
2005. for many international journals. He organized numerous events and served
as the General Chair, TPC Chair and other executive roles in many major
international events.
VOLUME 4, 2016 33
I. Cheikh et al.
MOHAMED SADIK received his Ph.D de- SÉBASTIEN ROY (Senior Member, IEEE) re-
gree in Electrical Engineering from the National ceived the B.Sc.A. and M.Sc. degrees in electri-
Polytechnic Institue of Lorraine (INPL) (Nancy, cal engineering from the Université Laval, Que-
France) in 1992. He is currently a full Professor bec, QC, Canada, in 1991 and 1993, respectively,
at the Department of Electrical Engineering at and the Ph.D. degree from Carleton University,
the National School of Electrical and Mechanical Ottawa, ON, Canada, in 2000. From 2000 to
Engineering (ENSEM) since 2008. He was chair 2002, he was a National Sciences and Engineering
of the Electrical Engineering Department and was Research Council of Canada (NSERC) Postdoc-
chair of the Research and cooperation department toral Fellow with Université Laval. In 2007, 2009,
at ENSEM from 2013 to 2019. He has founded and 2011, and 2018, he was an Invited Professor with
led the Networks, Embedded Systems and Telecom (NEST) Group Research the École Nationale Supérieure de Sciences Appliquées et de Technologie
(since 2011). Now he is chair of the Laboratory of Research in Engineering (ENSSAT), Lannion, France. In 2015, he was an Invited Professor with
(LRI Lab.). His previous research activities are part of the development of the Institut National de Sciences Appliquées (INSA), Rennes, France. From
autonomous biomedical systems. He is interested in codesign, modeling 2002 to 2012, he was an Assistant Professor, an Associate Professor since
and synthesis of Embedded Systems, autonomous/intelligent systems with 2005, and a Full Professor since 2010 with Université Laval. In 2012, he
applications to smart farming, smart health, smart environment, smart and joined the Electrical and Computer Engineering Department, Université de
gree energy, smart grid management, etc. His current research interests Sherbrooke, Sherbrooke, QC, Canada, where he is currently a Full Professor
include protocols design, ad hoc networking, networking game, pricing and was the Department Chair from 2016 to 2017, and where he is working
and networking neutrality, Image processing, machine learning and deep on system-oriented and implementation aspects of signal processing for
learning, Security in the cloud. He published several peer-reviewed papers advanced wireless communications. He is also a Co-Founder and a CoLeader
in reputed international journals and conferences and some book chapters. of the Center of Excellence in Integrated Intelligent Systems (CESIIUS),
He served on the technical program committees of many international Université de Sherbrooke. He has been active in industrial consulting and
conferences ( and as a reviewer of several journals). Pr. Sadik has co-founded was involved in the organization of several international conferences. Dr.
the International Symposium on Ubiquitous Networking (UNet). Roy received the Award for Post-Graduate Research Excellence from the
Canadian Institute for Telecommunications Research (CITR) in 2000. He
was a recipient of multiple teaching awards and is active in industrial
collaboration.
34 VOLUME 4, 2016

Multi-Layered Energy Efficiency in LoRa-WAN Networks A Tutorial

Uploaded by

Copyright:

Available Formats

Multi-Layered Energy Efficiency in LoRa-WAN Networks A Tutorial

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Multi-Layered Energy Efficiency in LoRa-WAN Networks A Tutorial

Uploaded by

Copyright:

Available Formats

This article has been accepted for publication in a future issue of this journal, but has not been

Multi-Layered Energy Efficiency in

I. INTRODUCTION Type Communications (eMTC) and Ultra-Reliable Low-

NET/MAC/PHY Energy Efficiency in LoRa-WAN Networks: A Tutorial

A systematic review of the real-time

An overview of the literature

LoRaWAN A systematic review of the literature

Using the impact of the confirmed traffic: The

Network performance analysis works of LoRa

(a) Energy Efficiency, Success and Coverage

A comprehensive review focusing on

A comprehensive review with a

No paper classification is done, the authors

Authors explore research challenges in terms of : Performance measures, current

Comprehensive tutorial covering

TABLE 1. Summary of recent related surveys.

FIGURE 2. Comparison between LPWANs Technologies [29].

- Narrowband IoT (NB-IoT) : Refers to an LPWAN radio

NET/MAC/PHY Energy Efficiency in LoRa-WAN Networks: A Tutorial

LoRaWAN Network Server

FIGURE 3. LoRaWAN architecture.

Standard LoRaWAN centered on chirp spread spectrum modulation (CSS) and

Security AED CCM 128 B. LORAWAN ARCHITECTURE

LoRa Modulation 2) LR-FHSS

NET/MAC/PHY Energy Efficiency in LoRa-WAN Networks: A Tutorial

of the payload gets determined by the first byte of the

lowing elements: MAC header (specifies the frame

signature of the payload);

NET/MAC/PHY Energy Efficiency in LoRa-WAN Networks: A Tutorial

Smart City Smart Agriculture

Smart Transport Smart Home

Smart Utilities Smart Grid

Urban Areas Rural Areas

FIGURE 7. LoRaWAN deployment opportunities.

[83]. H. LORAWAN SERVICES

munication module and its monthly fee. Class A

They are also required to listen to all channels at the same

LoRa networks have been widely deployed for a variety of Class B

measurement [20] [12]. Furthermore, paper [86] claims that Class C

NET/MAC/PHY Energy Efficiency in LoRa-WAN Networks: A Tutorial

Link Resource Channel Interference Key Third-Party

FIGURE 9. LoRaWAN challenges.

A. NETWORK SIDE The following is a collection of recent research on the above-

NET/MAC/PHY Energy Efficiency in LoRa-WAN Networks: A Tutorial

Adjusted Allocation Energy

Increases by 1% for a network

A solution based on machine It is always assumed

Running at the node, using With a highly loaded

A learning solution to adapt device The proposed scheme reduces

Totally ignores the

E-ADR is not suitable

The authors present the total

NET/MAC/PHY Energy Efficiency in LoRa-WAN Networks: A Tutorial

Adjusted Allocation Energy

Improves the packet error rate for

Maximizes the throughput of each It is restricted by

An adaptive solution to manage the Consumes more energy

General framework that considers

The scenario with the standard