15-09-2020

Internet of Things: IoT Reference Model Session Outline

1 IoT Architecture

2 IoT World Forum Reference Model

Dr. E.SURESH BABU
Assistant Professor 3 Physical Devices and Controllers Level

Computer Science and Engineering Department 4 Connectivity Level

National Institute of Technology, Warangal.
5 Edge Computing Level
Warangal, TS, India.
6 Data Accumulation Level

7 Data Abstraction Level

8 Application Level

Need of IoT Architecture

 Internet of Things is a very complex systems with many

“things”

IoT Architecture Devices, Gateways, and IoT Platforms  Various Sensors, Networking Technologies, Cloud Technology

Applications, Protocols, Hardware, Software etc. which needs a

overall plan

 IoT requires an Architecture

IoT Architecture IoT Architecture Standard

 Architecture is all about the protocol stack that provides  Architectures will define the standards and techniques for

 The relationship between the physical layer technologies designing and building IoT Ecosystem

and the applications.  Architectural standards and frameworks have emerged to

 The essence of IoT Architectures involves how the data is address the challenge of designing massive-scale IoT

transported, collected, analysed, and ultimately acted upon. networks.

 15-09-2020

Benefits of IoT Architecture Design Principles

 It provides the IT or network manager with a useful checklist for  Design principles that need to consider for an Ideal IoT Architecture.

evaluating the functionality and completeness of vendor  A Layering Approach in need to be considered

offerings.  To reduce the design complexity and to flexible implementation, the

concept of layering is introduced

 It provides guidance to developers as to which functions are
 IoT are organized as a Stack of Different Layers
needed in an IoT and how these functions work together.
 Stack for Constrained Devices, Stack for Gateways, Stack for IoT
 It can serve as a framework for standardization, promoting
Cloud Platforms
interoperability and cost reduction.

The IoT World Forum (IWF) Reference Model

 The IoT World Forum (IWF) is an Industry-sponsored Annual event

 Business, Government, and Academia will be part to promote the

IoT World Forum (IWF) Reference Model market adoption of IoT.

 Industry leaders including Cisco, IBM and Intel are part of this

event.

 This model serves as a common framework to help the industry

accelerate IoT deployments.

IWF Reference Model Characteristics of IWF Model

 The IWF model is mainly concerned to solve the broader issue  The IWF reference model is designed to have the following

 Developing The Applications, Characteristics.

1. Simplifies: It helps break down complex systems and each part

 Middleware, And
is understandable.
 Support Functions for an Enterprise-based IoT.
2. Clarifies: It provides additional information to identify levels of

the IoT and to establish common terminology.

 15-09-2020

Characteristics of IWF Model IoT World Forum Reference Model

3. Identifies: It identifies where specific types of processing is

optimized across different parts of the system.

4. Standardizes: It provides a first step in enabling vendors to

create IoT products that work with each other.

5. Organizes: It makes the IoT real and approachable, instead of

simply conceptual

IWF Reference Model IoT World Forum Reference Model

 IoT World Forum (IWF) is a seven-layer IoT architectural

reference model.

 Each of the seven layers is broken down into specific functions,

and security encompasses the entire model.

It provides a concise way of visualizing IoT from a technical

perspective.

Set of Levels IoT World Forum Reference Model

 The IoT Reference Model defines a set of levels with control

flowing from the Centre to the Edge

 Centre - Cloud Service Or A Dedicated Data Centre

 Edge- includes Sensors, Devices, Machines, and other types of

Intelligent End Nodes.

 Data travels up the stack, originating from the edge, and goes

northbound to the centre.

 15-09-2020

IT and OT Bridging IT and OT

 The key difference between IT and IoT is the data.

 The IT systems are mostly concerned with reliable and

continuous support of business applications such as email,

web, databases.

 The IoT is all about the data generated by sensors and how

that data is used.

IT and OT Role IoT in Process Layers

 The bottom of the stack is generally in the domain of OT.
People Process
 Sensors and Devices, Networking, Fog Computing and so on.
Application

 The top of the stack is in the IT area that includes Data Analysis IT Query
Based
Data at
Rest

Data Ingestion
 Things like the servers, databases, and applications, all of Cloud Infrastructure

Wide Area Connectivity

which run on a part of the network controlled by IT. Data Element Analysis
And Transformation Edge Computing
OT Event
Based
Data in
Motion
5G
Will have both

 OT and IT have generally been very independent and had little need Local Area Connectivity
(Lola, Sigfox, etc.)
Things, Devices, Sensors Edge
to even talk to each other.

Data Captured from Devices and Send to

Open IoT Framework Application as requested
 15-09-2020

Challenging Issue in IoT IT and OT Challenge

 In bottom, OT layers, the devices generate real-time data at their own
People Process rate.
Not capable of performing Real-time
Application
data analysis and feedback  Huge amount of data transiting the IoT network
Data Analysis
 Large volume of data available to the applications at the top layer (IT
Data Ingestion

Layer)
Edge Computing  Need to ingest that much data at the rate required.

ISSUE: Devices will  The IT and OT organizations need to work together for overall data
Edge
Generate data faster
management
than IT side can ingest it

Challenge: Minimizing IoT Data IT and OT Solution

 To meet this challenge

 Data has to be buffered or stored at certain points within the IoT

Minimize data before upload by extracting
and sending only the context = metadata Cloud
stack.

 Layering data management of IoT stack helps the top four layers handle

META data at their own speed.

DATA
 The real-time “data in motion” close to the edge has to be organized

and stored so that it becomes “data at rest” for the applications in

the IT tiers.

Roll of BIG DATA and AI in IoT Solution

People Process
People Process
Application
Application AI
AI
Data Analysis
Data Analysis
BIGDATA
BIGDATA Data Ingestion
Data Ingestion
Cloud Computing

AI
Edge Computing
Fog Computing Edge Computing

AI
Edge
Edge
 15-09-2020

Physical Devices

Physical Devices and Controller Layer

Physical Devices and Controllers Level Physical Devices & Device Controllers
 IoT “devices” are capable of:
 The first layer of the IoT Reference Model is  Analog to digital conversion, as required
 Generating data
 The physical devices and controllers layer.
 Being queried / controlled over-the-network

• Devices interact with physical things, such as sensors and

actuators that send and receive information.

 This Level-1 also control the multiple devices.

Capabilities of the Physical Devices Role and Size of the Physical Devices
 The capabilities of the devices may have  The primary function of these devices is generating data and

 Analog-to-digital and Digital-to-analog conversion, being capable of being queried and/or controlled over a

network.
 Data generation, and

 The size of the “things”

 The ability to be queried/controlled remotely.

 Range from almost microscopic sensors to giant machines

 The IoT Reference Model generally describes the level of
in a factory.
processing needed from Level 1 devices
 15-09-2020

Level-2 : Connectivity Level

Level-2 : Connectivity Level

Functions of Connectivity Level : (Communication

Connectivity : (Communication & Processing Units)
& Processing Units)
 Connectivity is in Level 2  Connectivity includes:

 Communicating with and between the Level-1 devices

 The functionality of this connectivity focuses on
 Reliable delivery across the network(s)
communications
 Implementation of various protocols

 Switching and routing

 Translation between protocols

 Security at the network level

 (Self Learning) Networking Analytics

Connectivity Level View Connectivity Level View

 Logical Point of View  Physical Point of View,

 This level enables communication between  This level consists of networking devices
• Physical devices (Level-1) and network (Level-2)
• Routers, Switches, Gateways, and Firewalls that are used
• Across networks (east-west)
to construct local and wide-area networks and provide
• Network (Level-2) and the Edge Computing Level (Level-3) for
Internet connectivity.
low-level information processing.
 15-09-2020

Role of Connectivity
 The most important function of this IoT layer is the reliable and

timely transmission of data.

• It provides communications and processing to be executed with

new and existing networks.

Level-3 : Edge Computing Level
• It enables devices to communicate with one another

• To communicate, via the upper logical levels, with application

platforms such as computers, remote control devices, and

smartphones.

Level-3 : Edge Computing Level Layer 3: Edge Computing Layer

 Edge computing is the role of Layer 3.

 Edge computing is often referred to as the “fog” layer.

 Many IoT Deployments,

 Massive amounts of data may be generated by a distributed

network of sensors.

Massive Amounts of Data Issue with Traditional IT

 For Instance
People Process

Application Not capable of performing  An Airplane can create multiple terabytes of data per hour.
Real-time data analysis
Data Analysis
and feedback  Storing all of the data permanently in central storage
Data Ingestion

accessible(Cloud) in big problem and send to the IoT

Edge Computing
applications
ISSUE: Devices will
Edge
Generate data faster
than IT side can ingest it
 15-09-2020

Need of Fog Computing Layer 3: Edge Computing Layer

 Level 1 devices do not include computing capabilities

themselves.

 Some computational activities could occur at Level 2, such as

protocol translation or application of network security policy.

IT Data Processing IoT Edge Computing Level

Computing
is a New
Computing
Data to Processing Paradigm
• Cloud
• Big Data
• Analytics
• Applications

Processing Data

Processing to Data

IoT Device Edge Node Fog Node Cloud

Data Data Data

Processing Processing Processing

© 2016 Cisco and/or its affiliates. All rights reserved. Cisco Public

The IoT Data Processing IoT Edge Computing Solutions

Computing
is a New
Data to Processing Computing  it is often desirable to do as much data processing close to the
Paradigm
• Cloud
• Big Data sensors as possible.
• Analytics
• Applications

Processing Data
 This Layer is to evaluate of data to see

 If it can be filtered or aggregated before being sent to a higher

Processing to Data

Cloud
layer.
IoT Device Edge Node Fog Node
Data Data Data
 It also allows for data to be reformatted or decoded, making

Processing Processing Processing additional processing by other systems easier.

© 2016 Cisco and/or its affiliates. All rights reserved. Cisco Public
 15-09-2020

Fog Computing Solution : The IoT Data Processing

Cloud

 Fog Computing and Fog Services are expected to be a Pass to Cloud

Data Pipeline Integrate with IT

Data Center
distinguishing characteristic of the IoT. Format for a Format for a
specific target specific target

 It handles massive numbers of individual smart objects Fog Node Store as

Time / Series Data
Transform to a
canonical format

that are interconnected with fog networking and

Edge Node

 Facilities for processing and storage resources close to the Communicate

with Device
Filter / Evaluate
Data

Device
edge devices in an IoT.

How can you use the IoT Data Via Fog Edge (Fog) Computing
7
Securely access
Cloud- based
Internet gateways/devices
applications fromanywhere  Mainly used to provide the Information understandable to the higher levels
6 Move data
programmatically to
your choice
of applications
Onpremise 5 Data packets
Controlwhich
applications data goes where,
4 View, monitor,
enforceownership,
manage
in realtime
security,privacy
 Include
Data Center

 Data filtering, Clean-up, Aggregation

Fog Nodes
3
Compute at edge
and fog layers for
 Packet content inspection
action,efficiency
2 Edge Nodes
Extract data
from anysource,  Combination of network and data level Analytics
any protocol, and 1 Automatically
normalize
provision via
cloud application
Devices  Event Generation
(Things)

©2017 Cisco and/or its affiliates. All rights reserved. Cisco Confidential

Edge (Fog) Computing Connectivity and Data Element Analysis

 Heterogeneous systems distributed across multiple

management domains represents an example of fog computing.

 Level 3 processing is performed on a packet-by-packet basis.

 Compute tasks such as packet inspection.

 The processing is limited, because there is only awareness of

data units—not “sessions” or “transactions.”

 15-09-2020

Introducing IoT “Edgeware” Role of Edge Computing

 The purpose of the edge computing level is

 To convert network data flows into information that is suitable

for storage and higher level processing.

 Processing Elements at this level may deal with

 High Volumes of Data and

 Perform Data Transformation Operations,

 Storage of much lower volumes of data.

Edge Computing Operations Edge Computing Operations

 Evaluation: Evaluating data for criteria as to whether it should  Distillation/Reduction: Reducing/summarizing data to

be processed at a higher level minimize the impact of data and traffic on the network

 Formatting: Reformatting data for consistent higher-level and higher-level processing systems

processing  Assessment: Determining whether data represents a threshold

 Expanding/Decoding: Handling cryptic data with additional or alert; this could include redirecting data to additional

context (such as the origin) destinations

Solutions to IoT Challenges

 It addresses various challenges raised by

 The activity of thousand or millions of smart devices,

including security, privacy, network capacity constraints, Level-4 : Data Accumulation Level
and latency requirements.

 The term fog computing is inspired by the fact that fog tends to

low to the ground, whereas clouds are high in the sky.

 15-09-2020

Level-4 :Data Accumulation (Storage) Data in Motion

Data in Motion Level-4 : Data Accumulation Level

 Data moving through a network is referred to as Data In Motion.  Networking systems are built to reliably move data. The data

 The rate and organization of the data in motion is determined by the is “in motion.”
devices generating the data.
 Data is moving through the network at the rate and

determined by the devices generating the data. The model is

event driven.

Level-4 : Data Accumulation Level Separation of Data

 This level marks a clear distinction

 The Requirements and Design Issues,

 Method of processing between lower-level (fog) computing

and upper-level (cloud computing.)

 15-09-2020

Data Accumulation (Storage) Level-4 : Data Accumulation Level

 Applications typically assume that data is “at rest” or unchanging

in memory or on disk.

 As Level 4 captures data and puts it at rest, it is now usable by

applications on a non-real-time basis.

 Applications access the data when necessary.

 This is a crucial step in bridging the differences between the real-

time networking world and the non-real-time application world.

Level-4 : Data Accumulation Level Operations Performed On Data Accumulation Level:

 At Level 4, Data Accumulation, data in motion is converted to  Converts data-in-motion to data-at-rest

data at rest.  Converts format from network packets to database relational tables.

 Filtered and Processed that are coming from the numerous  Achieves transition from event based to query based computing

devices at edge computing level

 Dramatically reduces data through filtering and selective storing

 stored in this level that will be accessible by higher levels.

Data Accumulation Level Determines

 Provides data is of interest to higher levels.  If data is of interest to higher levels: configured to serve the

 Maintains Persisted data. specific needs of a higher level.

 Type of storage is Needed.  If data must be persisted: Should data be kept on disk in a

 Verifies the data is organized properly non-volatile state or accumulated in memory.

 Checks the data must be recombined or recomputed  Type of storage needed: Does persistency require a file

system, big data system, or relational database?

 15-09-2020

Determines
 If data is organized properly: Is the data appropriately

organized for the required storage system?

 If data must be recombined or recomputed: Data might be Level-5 : Data Abstraction Level
combined, recomputed, or aggregated with previously stored

information may have come from non-IoT sources.

IoT World Forum Reference Model Role of Data Accumulation Level

 The Data Accumulation Level

 absorbs large quantities of data and places them in storage,

with little or no modifying to specific applications or groups of

applications.

 A number of different types of data in varying formats and from

heterogeneous processors may be coming up from the edge

computing level for storage.

Problems of Generating Data By Multiple Devices Solutions Provided by Data Abstraction

 Data may not land in the Same Data Storage.  Different kinds of data processing might be required.

 Too much data to put in one place.  Data storage for streaming data may be a big data system,

 Moving data into a database might consume too much such as Hadoop.

processing power  Storage for event data may be a relational database

management system (RDBMS) with faster query times.

 15-09-2020

The Legacy Application Compatibility Handling the Volume of Data

 The Legacy Application Compatibility

Sequence Pattern (use the existing DB and
schema)

Role of Data Accumulation Level Data Abstraction Level

 Levels 3 and 4 might separate “continuous streams of raw 1. Scalability
2. Interoperability: Enable Edgeware
data” from “data that represents an event.” and Applications from Different
Vendors
 Devices might be geographically separated, and processing

is optimized locally.

Data Abstraction (Aggregation & Access) Level-5 : Data Abstraction Level

 Information Integration
 The data abstraction level can aggregate and format this data
1. Creates schemas and views of data in the manner that applications want
in ways that make access by applications more.
2. Combines data from multiple sources, simplifying the application

3. Filtering, selecting, projecting, and reformatting the data to serve the client applications  The data abstraction functions are focused on interpretation
4. Reconciles differences in data shape, format, semantics, access protocol, and security
data that enable developing simpler, performance-enhanced

applications.

 manageable and efficient

 15-09-2020

Abstraction API between Data and Application Operation/Task on Data Abstraction Level
 Combining multiple data formats from different sources. This includes
integration multiple data formats.

 Perform necessary conversions to provide consistent semantics of

data across sources.

 Place formatted data in appropriate database.

 High-volume repetitive data may go into a big data system.

 Event data would be steered to a relational database management
system

Operation/Task on Data Abstraction Level

 Alerting higher-level applications that data is complete or had
accumulated to a defined threshold.

 Consolidating data into one place or providing access to multiple data

Level-6 : Application Level
stores through data virtualization.

 Protecting data with appropriate authentication and authorization.

 Normalizing or Denormalizing and indexing data to provide fast

application access.

Level-6 : Application Level Level-6 : Application Level

 In the application level,

 where information interpretation occurs.

 Software at this level interacts with Level 5 and data at

rest.
 15-09-2020

Application (Reporting, Analytics, Control) Types of Applications

 Applications vary based on vertical markets, the nature of

device data, and business needs.

 Some applications will focus on monitoring device data.

 Some will focus on controlling devices.

 Some will combine device and non-device data.

 Mobile applications that handle simple interactions

Types of Applications IoT Analytics Introduces New Complexities to Analytics

 Analytic applications that interpret data for business decisions

 Business intelligence reports, where the application is the BI

server

Key Issues:
 The velocity and volume of data may be huge
 In some cases, most of the data is unimportant

The Internet of Things and Analytics Analytics Using Both OT and IT Data

Key Points: IT – OT

 Interoperability

 Legacy

 Compatibility
 15-09-2020

Analytics Sequence Pattern Controls the IoT Devices

 This level contains any type of application that uses IoT input

or controls IoT devices.

 System management/control centre applications that control

the IoT system itself and don’t act on the data produced by it

Application Level Observation

 This level contains any type of application that uses IoT input or  If Levels 1-5 are architected properly, the amount of work
controls IoT devices.
required by Level 6 will be reduced.
 Provision should be available for streamlined operation that allows
 If Level 6 is designed properly, users will be able to do their
applications to bypass intermediate layers and interact directly with
Edge Computing Level or even Connectivity Level jobs better.

 Provision should be available for streamlined operation that allows

applications to bypass intermediate layers and interact directly with
Edge Computing Level or even Connectivity Level

IoT World Forum Reference Model

Collaboration and Processes Level

 15-09-2020

Communication and Collaboration Collaboration and Processes Level

 Communication and collaboration often

 Requires multiple steps. And it

 Usually transcends multiple applications.

 Level 7 represents a higher level than a single application.

Level-7 Collaboration and Processes Level Collaboration and Processes Level

 IoT need to include people and processes.  This may involve multiple applications and exchange of data
 IoT creates the information, is of little value unless it yields
and control information across the Internet or an enterprise
action, which often requires people and processes.
network.
 People must be able to communicate and collaborate to make
 Applications execute business logic to empower people.
an IoT useful.
 Multiple people use the same application for a range of
 People use applications and associated data for their specific

needs.
different purposes.

Example of IWF Reference Model

Example of IWF Reference Model

 15-09-2020

Summary of the IoT Reference Model

 The IWF views the IoT reference model as an industry-accepted
framework aimed at standardizing the concepts and terminology
associated with IoT
Edge
 The IWF model sets out the functionalities required and concerns that
Devices Fog
Devices must be addressed before the industry can realize the value of the
IoT

 This model is useful both for suppliers who develop functional

Network Cloud elements within the model and customers for developing their
Applications
Connectivity Infrastructure
requirements and evaluating vendor offerings.

Internet of Things Reference Model: Security

Security in the IoT

Internet of Things Reference Model: Security

 Discussions of security for each level and for the movement of data

between levels could fill a multitude of papers.

 For the purpose of the IoT Reference Model, security measures must:
Thank U
 Secure each device or system

 Provide security for all processes at each level

 Secure movement and communication between each level, whether

north- or south-bound