Performance Analysis of IEEE 802.

11ac DCF with

Hidden Nodes
Zheng Chang∗ , Olli Alanen∗ , Toni Huovinen∗ , Timo Nihtilä∗ , Eng Hwee Ong† , Jarkko Kneckt† and Tapani Ristaniemi‡
∗ Magister
Solutions Ltd, Hannikaisenkatu 41, FIN-40010 Jyväskylä, Finland
Email: {zheng.chang, olli.alanen, toni.huovinen, timo.nihtila}@magister.fi
† Nokia Research Center, Itämerenkatu 11-13,FIN-00180 Helsinki, Finland

Email: {eng-hwee.ong, jarkko.kneckt}@nokia.com

‡ Department of Mathematical Information Technology, University of Jyväskylä, FIN-40014 Jyväskylä, Finland

Email: {tapani.ristaniemi}@jyu.fi

Abstract—Recently, the IEEE 802.11 standard based Wireless model to study the performance of 802.11 DCF by assuming
Local Area Networks (WLAN) have become more popular and saturated traffic condition. Later on, [2] proposed an accurate
are widely deployed. It is anticipated that WLAN will play an unsaturated system analysis. The throughput analysis is shown
important rule in the future wireless communication systems
in order to provide several gigabits data rate. IEEE 802.11ac in [3] by taking into account a Hidden Node (HN) scenario,
is one of the ongoing WLAN standard aiming to support which is an important problem inherent to the basic access
very high throughput (VHT) with data rate of up to 6 Gbps scheme of DCF. With the ever increasing popularity of IEEE
below the 6 GHz band. In the development of IEEE 802.11ac 802.11 standard based WLAN, it is highly probable that a
standard, several new physical layer (PHY) and medium access station (STA) is in the coverage area of overlapping Basic
control layer (MAC) features are taken into consideration, such
as employing wider bandwidth in PHY and incrementing the Service Set (OBSS), which may result in a HN problem.
limits of frame aggregation in MAC. However, due to the newly Meanwhile, due to the high speed development of WLAN
introduced features, some traditional techniques used in previous technique, the performance of the most recent IEEE 802.11
standards could face some problems. This paper presents a per- system has drawn many interests. The IEEE 802.11ac is an
formance analysis of 802.11ac Distributed Coordination Function
ongoing next-generation WLAN standard which aims to offer
(DCF) in presence of hidden nodes in overlapping BSS (OBSS)
environment. The effectiveness of DCF in IEEE 802.11ac WLAN data rate of up to 6 Gbps [4]. Although the performance
when using different primary channels and different frequency of 802.11 DCF has been investigated [3] [5], new standard
bandwidth has also been discussed. Our results indicate that the introduces several new PHY and MAC features into WLAN,
traditional RTS/CTS handshake mechanism faces shortcomings e.g., usage of wider bandwidth from 40MHz in 802.11n to
and needs to be modified in order to support the newly defined
80 MHz or 160 MHz. This brings different impacts on the
802.11ac amendment.
Index Terms—WLAN, Distributed Coordination Function, performance of the DCF scheme. In this paper, we analyze
RTS/CTS, IEEE 802.11ac, Hidden Node the system performance of DCF with HN by taking into
account some of the new 802.11ac PHY and MAC features.
I. I NTRODUCTION One key challenge of 802.11ac specification formulation is the
indication and usage of different primary channels in different
The IEEE 802.11 Wireless Local Area Networks (WLAN) bandwidth, which have not been considered by current DCF
is widely known and used for its convenience and low costs. It researches. Therefore, we also present and analyze the results
can be anticipated that WLAN will play an important role in about the usage of different primary channels in OBSS.
the future wireless communication systems in order to provide The remainder of this paper is organized as follows: Section
data rate solutions of multi-gigabits transmission. So far, the II gives an overview of the 802.11 DCF as well as the
current IEEE 802.11n can offer data rate up to 600 Mbps new features of PHY and MAC in 802.11ac. In Section
and IEEE 802.11ac is one of the ongoing WLAN standards III, we use the existing mathematical throughput analysis
aiming to support Very High Throughput (VHT) with data model for analyzing 802.11ac DCF. Simulation results and
rate of up to 6 Gbps below 6 GHz band. In 1990’s, the IEEE performance analysis are presented in Section IV and Section
standardization group had defined the Distributed Coordination IV summarizes this paper.
Function (DCF) as the fundamental medium access method,
which is based on Carrier Sense Multiple Access with Colli- II. P RELIMINARY
sion Avoidance (CSMA/CA) technique. Since then researchers
A. IEEE 802.11 Distributed Coordination Function
have devoted a considerable amount of attentions on the
MAC performance of IEEE 802.11 systems. The performance IEEE 802.11 DCF is based on CSMA/CA, which adopts
analysis of IEEE 802.11 DCF is presented by Bianchi in [1]. In carrier sensing to avoid channel collisions. It employs two
this paper, the author proposed a bi-dimensional Markov chain techniques for the data transmission, the mandatory basic

978-1-4673-0990-5/12/$31.00 ©2012 IEEE

 Transmitter
access scheme (Fig. 1) and optional request to Request- BACK
RTS Data
To-Send/Clear-To-Send (RTS/CTS) mechanism (Fig. 2). The (BAR)

default mode is base access scheme, which is a two-way

Receiver
atomic exchange sequence that allows each STA lock out CTS ACK(BA)
the contention so that the atomic sequence is not interrupted
by other contending STAs. A STA with new packets for DIFS SIFS SIFS SIFS SIFS DIFS
transmitting will firstly monitor the channel. If the channel is Other device
sensed idle for an interval time exceeding the DCF Interframe NAV(RTS) Backoff

Space (DIFS), the STA may starts the packet transmission im- NAV(CTS)

mediately. Otherwise, the STA keeps monitoring the channel

activity, then enters the back-off status and randomly (also Figure 2: RTS/CTS access scheme
uniformly) generates a back-off time within a Contention
Window (CW) size before transmitting, e.g., in the range
[0, CW ] . The value of CW starts with a minimum value sensing mechanism is responsible for detecting the transmis-
CWmin , which we denote as W0 . Then the value doubles sion of other STAs by using Clear Channel Assessment (CCA)
after each unsuccessful transmission up to a maximum size function that resides in PHY. Hence, if another contending
CWmax . The relation between CWmin and CWmax could be STA is out of sensing range of sending STA, then collisions
defined as: CWmax = 2m CWmin , where m is the maximum could occur. Considering the situation in Fig. 3, there is data
increasing factor. The back-off timer is decreased by one transmission between STA1 and Access Point (AP)1. The
slot time if channel is sensed to be free in a slot time. If cycles are represented as the transmission range of STA3 and
transmission is detected on the channel, then the back-off timer STA1, respectively. When STA2 wants to send data to AP1,
is frozen and restarted only when channel becomes idle for it can detect the transmissions from STA1, and hence differs
more than a DIFS period. When the back-off timer reaches transmission. However, the distant STA3 is out of transmission
zero, the STA starts transmission. range of STA1. Thus, when STA1 transmits data to AP1, STA3
In 802.11ac, the sending STA will firstly send Block Ac- would not be able to detect the transmission and consider
knowledgment Request (BAR) after Short Interframe Space channel to be free. Therefore, collision will occur at AP1
(SIFS) period, then receiver responds with a Block Acknowl- when STA3 starts the transmission at the same time. In such
edgment (BA) frame. If BA is not received by the sending case, STA3 is called Hidden Node (HN) with respect to the
STA, it will start its back-off procedure and double its current communication between STA1 and AP1.
CW unless CW = CWmax . If BA is received or maximum HN problem can’t be solved by the basic backoff rules in
retry limits is reached, the CW is always reset. MAC since STA fails to sense other existing transmissions.
The RTS/CTS scheme, which relies on virtual carrier sensing
Transmitter Backoff Data BAR mechanism, is one typical solution for HN problem. However,
the exchange of RTS and CTS frames brings longer MAC
Receiver ACK(BA) overhead and costs more radio resource. The effectiveness of
RTS/CTS is controversial, and it has been investigated in some
DIFS SIFS SIFS DIFS
papers with existing IEEE standard [5] [6].
Figure 1: Basic access scheme

In the RTS/CTS mechanism, besides following the above STA2

mentioned basic access scheme, the sending STA will send AP1
STA3

a special RTS frame after medium is sensed to be free for STA1

a DIFS period. When the receiver receives the RTS frame,

after a SIFS period it will respond with CTS frame. The Figure 3: Hidden node problem
transmission is started by sending STA only if the CTS frame
have been received correctly. During the RTS/CTS exchange
period, the other contending STAs also read the information of C. Overview of Key PHY Features and MAC Enhancement of
RTS/CTS frames and update their Network Allocation Vector 802.11ac
(NAV) containing the information of which period the medium
remains captured. IEEE 802.11ac, which aims to provide VHT below the 6
GHz band, is currently under development. It could be viewed
B. Hidden Node Problems as an extension of the existing 802.11n standard, where basic
Carrier sensing mechanism or listen-before-talk scheme is notions of Multiple Input Multiple Output (MIMO) and wider
critical for collision avoidance due to the inherent property channel bandwidth are enhanced generally. An overview of
of the DCF scheme. 802.11 employs both physical carrier the key PHY features and MAC enhancements of 802.11ac
sensing and virtual carrier sensing mechanisms. The carrier are introduced in the following.
 1) Mandatory and Optional PHY Features: The key feature where n is the number of contending STAs, W0 = CWmin
that makes 802.11ac different from 802.11n in PHY is the and m is the maximum increasing factor. The transmission
support for 80 MHz or 160 MHz channel bandwidths. Usage probability τ and collision probability p can be calculated by
of 80 MHz channel can approximately double the data rate solving the nonlinear equations of (1) and (2) numerically
as compared to 802.11n where 40 MHz is the largest channel using fixed point iteration technique. It can be proved that
bandwidth. As a result, only one spatial stream is mandatory the system has unique solutions [1]. The normalized system
in 802.11ac instead of one or two spatial streams as specified throughput S, which is defined as the ratio of the average
in 802.11n. As an optional feature, the support of 160 MHz number of successfully transmitted bits in a slot time over the
channel is also defined in 802.11ac for another two-fold average slot time, can be calculated as :
increase in data rate over 80 Mhz channel bandwidth which Ptr Ps E[P ]
is mandatory. Another feature of 802.11ac is that 256QAM is S= , (3)
T
added as an optional modulation scheme in order to support
peak data rates of close to 7 Gbps transmission while 64QAM where Ptr is the probability that there is at least one
is the highest modulation scheme specified in 802.11n. transmission is occurred on the channel in the considered
slot time, Ps is the probability that the transmission occurred
In order to support wider channel bandwidths, 802.11ac
is successful, and E[P ] is the average payload size. Since
defines its channelization for 20, 40, 80, and 160 MHz
there are n contending STAs on the channel and each of them
channels as shown in Fig. 4. For example, a 40 MHz band
transmits with τ , so we have Ptr = 1 − (1 − τ )n .
is formed by two contiguous 20 MHz bands, and 80 MHz
If we denote p = 1 − p as the probability that transmission
transmission band is formed by two contiguous 40 MHz bands,
is successful without collision. Then Ps can be expressed as:
where one of the 20 MHz bands is the primary channel and
the others are secondary channels. However, unlike only one nτ p nτ (1 − τ )n−1
20 MHz channel is specified in 802.11a, how to utilize and Ps = = . (4)
Ptr 1 − (1 − τ )n
support the multi-channel are still critical for the development
of 802.11ac specifications. The average slot time T can be calculated as [1]:

T = (1 − Ptr )σ + Ptr Ps Ts + Ptr (1 − Ps )Tc , (5)

where σ is the duration of an empty slot time, Ts and Tc are
the average times that channel is busy because of successful
transmission and collisions respectively. For the basic access
mechanism, Ts and Tc are expressed as:

Figure 4: Channelization in the draft IEEE 802.11ac standard Tcb = Tsb = Tdata−ba + DIF S (6)
For the RTS/CTS scheme,
2) MAC Enhancements: 802.11n introduces two kinds of
Tcrts = Tphy + Trts + DIF S,
frame aggregations comprising Aggregated MAC Protocol
Data Unit (A-MPDU) and Aggregated MAC Service Data Unit Tsrts= 2Tphy + Trts + 2SIF S + Tcts + Tdata−ba + DIF S,
(A-MSDU) to enhance its MAC efficiency. It is also possible to (7)
combine both which is referred as hybrid A-MSDU/A-MPDU where Trts and Tcts are the transmission time for RTS and
aggregation hereinafter. Due to multiple channels are defined, CTS frame respectively. Tdata−ba is the time for transmitting
the key MAC enhancements of 802.11ac are centered around data and BAR frames as well as receiving BA frame. We
its capability of multi-channel operations. More details of the assume that collision occurs only to the RTS frame, and
MAC layer enhancement as well as PHY features could be propagation delay is not taken into account accordingly. We
found in [7]. have
Tdata−ba = 3Tphy + 2SIF S + Tdata + TBAR + TBA , (8)
III. S YSTEM A NALYSIS
We recall the Markov chain model in [1]. STA starts
transmission in a generic time slot with probability τ , and Tdata = Tphy + Tsym Nsym , (9)
the transmission suffers from the collision with probability p.
where Tsym is the tranmission time for a symbol and Nsym
We assume saturated traffic condition, hence τ and p can be
is the number of symbols.
expressed as [1] :
2(1 − 2p) IV. P ERFORMANCE A NALYSIS
τ= , (1) A. Simulation Description
(1 − 2p)(W0 + 1) + pW0 (1 − (2p)m )
In this section, we investigate the performance of DCF
p = 1 − (1 − τ )n−1 , (2) with basic access scheme and RTS/CTS scheme under various
 Table I: Simulation parameters Fig. 5b. The main target for this simulation is to investigate
Name
Rdata
Description
Data rates
802.11ac
var.
the effectiveness of newly defined 80 MHz/160 MHz channels
Rcontrol RTS CTS ACK rates 6 Mbps as well as 256QAM modulation scheme. Although 160 MHz
SIFS SIFS duration 16 µs
DIFS DIFS duration 34 µs channel is an optional PHY feature for the ongoing 802.11ac
Lrts RTS frame size 20 bytes
Lcts CTS frame size 14 bytes standardisation, the effectiveness of DCF in 160 MHz channel
Lbar MAC compressed BAR frame size 24 bytes
Lba MAC compressed BA frame size 32 bytes remains relevant. Here, the usage of 80 MHz and 160 MHz
LM AChdr MAC overhead 34 bytes
CWmin,max Contention window size {15,127}
have almost the same performance in error-prone channel.
NAM P DU
NAM SDU
Number of aggregated MPDU
Number of aggregated MSDU
1/10(for simulation)
1/2(for simulation)
Similarly, the throughput of the basic access scheme outper-
LSER
LT AIL
Length of service bits
Length of tail bits
16 bits
6 bits
forms RTS/CTS mechanism.
NES Number of encode stream 2 (for simulation)
Tphy VHT-PHY and legacy preamble and header time 68.8 µs

40 MHz and 80 MHz channel 80 MHz and 160 MHz channel

60 10 STAs, 40MHz, basic DCF, simulations 50
10 STAs, 80MHz, basic DCF, simulations
10 STAs, 40MHz, with RTS/CTS, simulations
10 STAs, 80MHz, with RTS/CTS, simulations
50 STAs, 40MHz, basic DCF, simulations 45
simulation scenarios. To evaluate the performance of DCF 50
50 STAs, 40MHz, with RTS/CTS, simulations
50 STAs, 80MHz, basic DCF, simulations
50 STAs, 80MHz, with RTS/CTS, simulations
10 STAs, 80MHz, basic DCF, simulations
within 802.11ac environment, we propose different scenarios 10 STAs, 80MHz, with RTS/CTS, simulations
40 10 STAs, 160MHz, basic DCF, simulations
10 STAs, 160MHz, with RTS/CTS, simulations
with different data rates. At first we validate our numeri- 50 STAs, 80MHz, basic DCF, simulations
35 50 STAs, 160MHz, basic DCF, simulations

Throughput (Mbps)

Throughput (Mbps)
40 50 STAs, 80MHz, with RTS/CTS, simulations
50 STAs, 160MHz, with RTS/CTS, simulations
cal simulation results by considering single AP and various 10 STAs, 40MHz, basic DCF, theortical
50 STAs, 40MHz, basic DCF, theortical 30

numbers of STAs. The results also show the effectiveness of 30 25

DCF scheme in contending node scenario with wider channel 20

bandwidth. The we present the OBSS scenario including HN 20
15
to examine the performance of DCF and the usage of different
primary channels in multiple APs. In all these simulations, we 10
10

set Transmit Opportunity (TXOP) limit to be zero. Channel 5

propagation is modeled by using the IEEE TGac specifications 0

200 400 600 800 1000 1200 1400 1600
0
200 400 600 800 1000 1200 1400 1600
[4]. The channel bandwidth varies from 40 MHz to 160 MHz. Packet size(Bytes) Packet size(Bytes)
Note that if 40 MHz is used, the transmission is based on (a) 40 MHz and 80 MHz channels (b) 80 MHz and 160 MHz channels
the 802.11n HT-mixed format while others are based on the
802.11ac VHT format [7]. Figure 5: Error-prone channel

B. Simulation Results 2) OBSS with hidden node case: One of our main goal is
The parameters that are used in the simulation are based on to investigate the effectiveness of current DCF in OBSS with
the draft IEEE 802.11ac standard as shown in Table I. different usages of primary channels. One simplified scenario
1) Single AP with varying contending nodes: We show the is shown in Fig. 6:
effectiveness of DCF by taking into account a single AP and
different numbers of STAs. For validating our simulator, we
AP1
also show the numerical results for the basic access scheme. STA1 STA2
AP2
Fig. 5a plots the uplink throughput of whole BSS against the
different STA packet sizes in an error-prone channel, where BSS1’s primary

collision is the only reason that causes packet error. The goal BSS1’s channel
BSS2’s Primary

is to investigate the performance of DCF with 80 MHz channel BSS2’s channel

bandwidth. To fulfill this goal, we compare it with 40 MHz Figure 6: OBSS with hidden node case
channel with 64QAM modulation scheme, which provides
the highest data rate for 802.11n. The data rate considered In Fig. 6, we consider AP1-STA1 pair as BSS1 and AP2-
in simulations is 270 Mbps for both 40 MHz and 80 MHz STA2 pair as BSS2, and note that STA1 and STA2 are hidden
channels. We do not use aggregation scheme here. First, we from each other due to limitation of sensing range. Fig. 6
can see that the theoretical results match the simulation results shows that one STA is associating with one AP, but it can
very well. Generally, we observe that, the throughput of basic be extended readily to multiple STAs associating with same
access scheme outperforms the RTS/CTS scheme, especially AP. Since we assume that AP2 can still hear the transmission
with fewer STAs. The difference is not so obvious when more of BSS1, the transmission between STA2 and AP2 could
STAs are considered. We notice that the usage of 80 MHz be interfered by STA1’s transmission. We consider uplink
channel provides only a little improvement (about 2 Mbps) transmissions only, and hence APs do not send any data to its
comparing to the usage of 40 MHz channel in the case of 50 STA but only control frames, such as CTS and BACK. The
STAs. However, in the case of 10 STAs, the usage of 40 MHz number of A-MPDU and A-MSDU are 10 and 2, respectively.
can provide a larger throughput of up to 10 Mbps. Fig. 7 shows the throughput performance of both BSSs
We also compare the throughput performance between 80 when 40 MHz and 80 MHz channels are used. Both BSSs
MHz and 160 MHz channels with data rates of 270 Mbps in are using the same primary channel and an error-prone channel
model is considered. The goal is to study how collision caused
40 MHz/270Mbps, 80 MHz/540Mbps 40 MHz/270Mbps, 80 MHz/540Mbps
by HN problem affects the transmission of OBSS. Generally, 250 250

we can see that BSS2 has worse throughput performance

due to the HN problem. This is due to the fact that the 200 200

Uplink throughput(Mbps)

Uplink throughput(Mbps)
transmissions from STA2 to AP2 could be interfered by the
transmissions of BSS1 but not vice versa. The RTS/CTS 150 150

scheme actually degrades the performance of BSS1 and pre- 40MHz,with RTS/CTS, BSS1 40MHz,with RTS/CTS, BSS1
100 100
serve the transmissions of BSS2. In Fig. 7, we notice that 80MHz,with RTS/CTS, BSS2 80MHz,with RTS/CTS, BSS2
40MHz, Basic DCF, BSS1 40MHz, Basic DCF, BSS1
how frequent collisions happen during transmission since the 80MHz, Basic DCF, BSS2 80MHz, Basic DCF, BSS2
50 50
throughput of BSS2 is almost zero when basic access scheme
is used. Although the RTS/CTS scheme improve the through-
0 0
put of BSS2 by up to 40 Mbps due to its ability to mitigate 0 500 1000
Packet size(Bytes)
1500 0 500 1000
Packet size(Bytes)
1500

HN problem, such effect is limited and the usage of the

(a) 40Hz/270Mbps, 80MHz/540Mbps (b) 40Hz/270Mbps, 80MHz/540Mbps
RTS/CTS scheme could degrade the throughput performance
of BSS1. The usage of 80 MHz channel can offer throughput Figure 8: Different primary channel
performance of up to 25 Mbps as compared to 40 MHz channel
for BSS1, but not for BSS2.
current DCF cannot fully preserve the throughput performance
40 MHz and 80 MHz channel, 270 Mbps of BSS2 as well as effectively utilizing the newly defined
250
802.11ac amendments.

200
V. C ONCLUSION AND F UTURE
Uplink throughput(Mbps)

In this work, we have studied the effect of current DCF

150
access mechanism in 802.11ac scenarios. The target for gigabit
40MHz,with RTS/CTS, BSS1 transmission and the support for multi-channel operations
40 MHz,with RTS/CTS, BSS2
40MHz, Basic DCF, BSS1
bring challenges for the emerging 802.11ac standardization.
100
40MHz, Basic DCF, BSS2 The impact of using the RTS/CTS and basic access schemex in
80MHz,with RTS/CTS, BSS1
80 MHz,with RTS/CTS, BSS2
VHT WLAN has been examined in the context of an OBSS in
50 80MHz, Basic DCF, BSS1 this work. We have concluded that the throughput performance
80MHz, Basic DCF, BSS2
of overall VHT WLAN system suffers from the drawbacks
0
of current DCF scheme. The benefits of wider channel band-
0 500 1000 1500
packet size(Bytes) width, different primary channel and higher order modulation
scheme can’t be utilized ultimately without enhancement of
Figure 7: Same primary channel RTS/CTS scheme, e.g. in [8]. For future work, we planned
to investigate an effective RTS/CTS scheme that can fully
Since the question of how to effectively support for multi- support multi-channel feature and preserve transmissions in
channels is critical for future WLAN research, we extend our the upcoming IEEE 802.11ac standard .
work to the study of usage of different primary channels in R EFERENCES
OBSS. We assume BSS1 uses only 40 MHz channel, while
[1] G. Bianchi, "Performance Analysis of the IEEE 802.11 Distributed Coor-
BSS2 is using 80 MHz channel and their primary channels are dination Function," IEEE Journal on Selected Areas in Communications
different. The data rates for BSS1 and BSS2 are 270 Mbps and , Vol.18, No.3, pp.535-547, Mar. 2000.
540 Mbps respectively. [2] D. Malone, K. Duffy and D. Leith, "Modeling the 802.11 Distributed
Coordination Function in Non-saturated Heterogeneous Conditions,"
The considered scenario in Fig. 8a is that BSS2 uses IEEE/ACM Transaction on Networking, Vol.15, No.1, pp.159-172, Jan.
secondary channel of BSS1 as its primary channel, while in 2007.
Fig. 8b, the primary channel of BSS2 does not lie on the [3] O. Ekici and A. Yongacoglu, "IEEE 802.11a Throughput Performance
with Hidden Nodes," IEEE Communications Letters, Vol.12, No.6,
channel bandwidth scale of BSS1. From Fig. 8, we see that the pp.465-467, Jun.2008.
throughput performance is similar to the one in Fig. 7 though [4] IEEE P802.11ac. Specification Framework for TGac.
the data rate is doubled for BSS2. The reason is that the current [5] P. Chatzimisions, et al, "Effectiveness of RTS/CTS handshake in IEEE
802.11a Wireless LANs," Electronics Letters, Vol 40, No.14, July 2004.
RTS/CTS scheme makes BSS2 fail to know the transmission [6] K. Xu, et al, "How effective is the IEEE 802.11 RTS/CTS handshake
of BSS1. Therefore, although BSS2 is using different primary in ad hoc networks," Proc. of IEEE Globalcom 2002, Taipei, Taiwan,
channel and bandwidth for transmission, it still suffers from Nov. 2002
[7] E. H. Ong, J. Kneckt, O. Alanen, Z. Chang, T. Huovinen, and T. Nihtila,
the transmission of BSS1. In Fig. 8b, we can notice that if "IEEE 802.11ac : Enhancements for Very High Throughput WLANs ,"
the primary channel of BSS2 does not lie on the channel Proc. of IEEE PIMRC 2011, Toronto, Canada, Sep. 2011.
bandwidth of BSS2, the throughputs of both BSSs become [8] M. Park, "IEEE 802.11ac : Dynamic Bandwidth Channel Access ," Proc.
of IEEE ICC 2011, Kyoto, Japan, Jun. 2011.
better (around 10 Mbps comparing to Fig. 8a) for the case
with RTS/CTS support. Therefore, we can conclude that the