IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS, VOL. 18, NO.

3, MARCH 2000 535

Performance Analysis of the IEEE 802.11 Distributed

Coordination Function
Giuseppe Bianchi

Abstract—Recently, the IEEE has standardized the 802.11 pro- mission of an ACK is required since, in the wireless medium, a
tocol for Wireless Local Area Networks. The primary medium ac- transmitter cannot determine if a packet is successfully received
cess control (MAC) technique of 802.11 is called distributed coor- by listening to its own transmission.
dination function (DCF). DCF is a carrier sense multiple access
with collision avoidance (CSMA/CA) scheme with binary slotted In addition to the basic access, an optional four way hand-
exponential backoff. This paper provides a simple, but nevertheless shaking technique, known as request-to-send/clear-to-send
extremely accurate, analytical model to compute the 802.11 DCF (RTS/CTS) mechanism has been standardized. Before transmit-
throughput, in the assumption of finite number of terminals and ting a packet, a station operating in RTS/CTS mode “reserves”
ideal channel conditions. The proposed analysis applies to both the the channel by sending a special Request-To-Send short frame.
packet transmission schemes employed by DCF, namely, the basic
access and the RTS/CTS access mechanisms. In addition, it also ap- The destination station acknowledges the receipt of an RTS
plies to a combination of the two schemes, in which packets longer frame by sending back a Clear-To-Send frame, after which
than a given threshold are transmitted according to the RTS/CTS normal packet transmission and ACK response occurs. Since
mechanism. By means of the proposed model, in this paper we pro- collision may occur only on the RTS frame, and it is detected
vide an extensive throughput performance evaluation of both ac- by the lack of CTS response, the RTS/CTS mechanism allows
cess mechanisms of the 802.11 protocol.
to increase the system performance by reducing the duration
Index Terms—802.11, collision avoidance, CSMA, performance of a collision when long messages are transmitted. As an
evaluation. important side effect, the RTS/CTS scheme designed in the
802.11 protocol is suited to combat the so-called problem of
I. INTRODUCTION Hidden Terminals [4], which occurs when pairs of mobile
stations result to be unable to hear each other. This problem

I N recent years, much interest has been involved in the

design of wireless networks for local area communication
[1], [2]. Study group 802.11 was formed under IEEE Project
has been specifically considered in [5] and in [6], which, in
addition, studies the phenomenon of packet capture.
In this paper, we concentrate on the performance evaluation
802 to recommend an international standard for Wireless Local
of the DCF scheme, in the assumption of ideal channel con-
Area Networks (WLAN’s). The final version of the standard
ditions and finite number of terminals. In the literature, perfor-
has recently appeared [3], and provides detailed medium access
mance evaluation of 802.11 has been carried out either by means
control (MAC) and physical layer (PHY) specification for
of simulation [7], [8] or by means of analytical models with sim-
WLAN’s.
plified backoff rule assumptions. In particular, constant or geo-
In the 802.11 protocol, the fundamental mechanism to access
metrically distributed backoff window has been used in [5], [9],
the medium is called distributed coordination function (DCF).
[10] while [11] has considered an exponential backoff limited to
This is a random access scheme, based on the carrier sense mul-
two stages (maximum window size equal to twice the minimum
tiple access with collision avoidance (CSMA/CA) protocol. Re-
size) by employing a two dimensional Markov chain analysis.
transmission of collided packets is managed according to bi-
In this paper, which revises and substantially extends [12], we
nary exponential backoff rules. The standard also defines an op-
succeed in providing an extremely simple model that accounts
tional point coordination function (PCF), which is a centralized
for all the exponential backoff protocol details, and allows to
MAC protocol able to support collision free and time bounded
compute the saturation (asymptotic) throughput performance of
services. In this paper we limit our investigation to the DCF
DCF for both standardized access mechanisms (and also for any
scheme.
combination of the two methods). The key approximation that
DCF describes two techniques to employ for packet transmis-
enables our model is the assumption of constant and indepen-
sion. The default scheme is a two-way handshaking technique
dent collision probability of a packet transmitted by each station,
called basic access mechanism. This mechanism is character-
regardless of the number of retransmissions already suffered. As
ized by the immediate transmission of a positive acknowledge-
proven by comparison with simulation, this assumption leads to
ment (ACK) by the destination station, upon successful recep-
extremely accurate (practically exact) results, especially when
tion of a packet transmitted by the sender station. Explicit trans-
the number of stations in the wireless LAN is fairly large (say
greater than ten).
Manuscript received November 1998; revised July 25, 1999. this work was The paper is outlined as follows. In Section II we briefly re-
supported in part by CNR and MURST, Italy. view both basic access and RTS/CTS mechanisms of the DCF.
The author is with the Universitá di Palermo, Dipartimento di Ingegneria Elet-
trica, Viale delle Scienza, 90128 Palermo, Italy (e-mail: bianchi@elet.polimi.it). In Section III we define the concept of Saturation Throughput,
Publisher Item Identifier S 0733-8716(00)01290-7. and in Section IV we provide an analytical technique to com-
0733–8716/00$10.00 © 2000 IEEE
Authorized licensed use limited to: Indian Institute of Technology Indore. Downloaded on February 06,2024 at 09:12:24 UTC from IEEE Xplore. Restrictions apply.
536 IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS, VOL. 18, NO. 3, MARCH 2000

pute this performance figure. Section V validates the accuracy TABLE I

of the model by comparing the analytical results with that ob- SLOT TIME, MINIMUM, AND MAXIMUM
CONTENTION WINDOW VALUES FOR THE THREE PHY SPECIFIED BY THE
tained by means of simulation. Additional considerations on the 802.11 STANDARD: FREQUENCY HOPPING SPREAD SPECTRUM (FHSS), DIRECT
maximum throughput theoretically achievable are carried out in SEQUENCE SPREAD SPECTRUM (DSSS), AND INFRARED (IR)
Section VI. Finally, the performance evaluation of both DCF ac-
cess schemes is carried out in Section VII. Concluding remarks
are given in Section VIII.

II. 802.11 DISTRIBUTED COORDINATION FUNCTION

This section briefly summarizes the DCF as standardized by sion, station B waits for a DIFS and then chooses a backoff time
the 802.11 protocol. For a more complete and detailed presen- equal to 8, before transmitting the next packet. We assume that
tation, refer to the 802.11 standard [3]. the first packet of station A arrives at the time indicated with an
A station with a new packet to transmit monitors the channel arrow in the figure. After a DIFS, the packet is transmitted. Note
activity. If the channel is idle for a period of time equal to a dis- that the transmission of packet A occurs in the middle of the Slot
tributed interframe space (DIFS), the station transmits. Other- Time corresponding to a backoff value, for station B, equal to 5.
wise, if the channel is sensed busy (either immediately or during As a consequence of the channel sensed busy, the backoff time
the DIFS), the station persists to monitor the channel until it is is frozen to its value 5, and the backoff counter decrements again
measured idle for a DIFS. At this point, the station generates a only when the channel is sensed idle for a DIFS.
random backoff interval before transmitting (this is the Colli- Since the CSMA/CA does not rely on the capability of the sta-
sion Avoidance feature of the protocol), to minimize the prob- tions to detect a collision by hearing their own transmission, an
ability of collision with packets being transmitted by other sta- ACK is transmitted by the destination station to signal the suc-
tions. In addition, to avoid channel capture, a station must wait cessful packet reception. The ACK is immediately transmitted
a random backoff time between two consecutive new packet at the end of the packet, after a period of time called short inter-
transmissions, even if the medium is sensed idle in the DIFS frame space (SIFS). As the SIFS (plus the propagation delay) is
time.1 shorter than a DIFS, no other station is able to detect the channel
For efficiency reasons, DCF employs a discrete-time backoff idle for a DIFS until the end of the ACK. If the transmitting sta-
scale. The time immediately following an idle DIFS is slotted, tion does not receive the ACK within a specified ACK_Timeout,
and a station is allowed to transmit only at the beginning of each or it detects the transmission of a different packet on the channel,
slot time. The slot time size, , is set equal to the time needed it reschedules the packet transmission according to the given
at any station to detect the transmission of a packet from any backoff rules.
other station. As shown in Table I, it depends on the physical
The above described two-way handshaking technique for the
layer, and it accounts for the propagation delay, for the time
packet transmission is called basic access mechanism. DCF de-
needed to switch from the receiving to the transmitting state
fines an additional four-way handshaking technique to be op-
(RX_TX_Turnaround_Time), and for the time to signal to the
tionally used for a packet transmission. This mechanism, known
MAC layer the state of the channel (busy detect time).
with the name RTS/CTS, is shown in Fig. 2. A station that wants
DCF adopts an exponential backoff scheme. At each packet
to transmit a packet, waits until the channel is sensed idle for a
transmission, the backoff time is uniformly chosen in the range
DIFS, follows the backoff rules explained above, and then, in-
. The value is called contention window, and de-
stead of the packet, preliminarily transmits a special short frame
pends on the number of transmissions failed for the packet. At
called request to send (RTS). When the receiving station detects
the first transmission attempt, is set equal to a value
an RTS frame, it responds, after a SIFS, with a clear to send
called minimum contention window. After each unsuccessful
(CTS) frame. The transmitting station is allowed to transmit its
transmission, is doubled, up to a maximum value
packet only if the CTS frame is correctly received.
. The values and reported in the
final version of the standard [3] are PHY-specific and are sum- The frames RTS and CTS carry the information of the length
marized in Table I. of the packet to be transmitted. This information can be read
The backoff time counter is decremented as long as the by any listening station, which is then able to update a network
channel is sensed idle, “frozen” when a transmission is detected allocation vector (NAV) containing the information of the period
on the channel, and reactivated when the channel is sensed idle of time in which the channel will remain busy. Therefore, when
again for more than a DIFS. The station transmits when the a station is hidden from either the transmitting or the receiving
backoff time reaches zero. station, by detecting just one frame among the RTS and CTS
Fig. 1 illustrates this operation. Two stations A and B share frames, it can suitably delay further transmission, and thus avoid
the same wireless channel. At the end of the packet transmis- collision.
The RTS/CTS mechanism is very effective in terms of system
1As an exception to this rule, the protocol provides a fragmentation mecha- performance, especially when large packets are considered, as
nism, which allows the MAC to split an MSDU (the packet delivered to the MAC it reduces the length of the frames involved in the contention
by the higher layers) into more MPDUs (packets delivered by the MAC to the process. In fact, in the assumption of perfect channel sensing
PHY layer), if the MSDU size exceeds the maximum MPDU payload size. The
different fragments are then transmitted in sequence, with only a SIFS between by every station, collision may occur only when two (or more)
them, so that only the first fragment must contend for the channel access. packets are transmitted within the same slot time. If both trans-
Authorized licensed use limited to: Indian Institute of Technology Indore. Downloaded on February 06,2024 at 09:12:24 UTC from IEEE Xplore. Restrictions apply.
BIANCHI: PERFORMANCE ANALYSIS OF THE IEEE 802.11 DCF 537

Fig. 1. Example of basic access mechanism.

Fig. 2. RTS/CTS Access Mechanism.
mitting stations employ the RTS/CTS mechanism, collision oc-
curs only on the RTS frames, and it is early detected by the
transmitting stations by the lack of CTS responses. A quanti-
tative analysis will be carried out in Section VII.

III. MAXIMUM AND SATURATION THROUGHPUT PERFORMANCE

In this paper we concentrate on the “Saturation Throughput”.
This is a fundamental performance figure defined as the limit
reached by the system throughput as the offered load increases,
and represents the maximum load that the system can carry in
stable conditions.
It is well known that several random access schemes exhibit
an unstable behavior. In particular, as the offered load increases,
the throughput grows up to a maximum value, referred to as
“maximum throughput.” However, further increases of the
offered load lead to an eventually significant decrease in the
system throughput. This results in the practical impossibility to
operate the random access scheme at its maximum throughput Fig. 3. Measured Throughput with slowly increasing offered load.
for a “long” period of time, and thus in the practical mean-
ingless of the maximum throughput as performance figure the example considered), but ultimately it decreases and stabi-
for the access scheme. The mathematical formulation and lizes to the saturation value. Queue build-up is observed in such
interpretation of this instability problem is the object of a wide a condition.
and general discussion in [13].
Indeed, the 802.11 protocol is known to exhibits some form IV. THROUGHPUT ANALYSIS
of instability (see, e.g., [5], and [11]). To visualize the unstable
behaviour of 802.11, in Fig. 3 we have run simulations in which The core contribution of this paper is the analytical evalu-
the offered load linearly increases with the simulation time. The ation of the saturation throughput, in the assumption of ideal
general simulation model and parameters employed are summa- channel conditions (i.e., no hidden terminals and capture [6]).
rized in Section V. The results reported in the figure are obtained In the analysis, we assume a fixed number of stations, each al-
with 20 stations. The straight line represents the ideal offered ways having a packet available for transmission. In other words,
load, normalized with respect of the channel capacity. The sim- we operate in saturation conditions, i.e., the transmission queue
ulated offered load has been generated according to a Poisson of each station is assumed to be always nonempty.
arrival process of fixed size packets (payload equal to 8184 bits), The analysis is divided into two distinct parts. First, we study
where the arrival rate has been varied throughout the simulation the behavior of a single station with a Markov model, and we
to match the ideal offered load. The figure reports both offered obtain the stationary probability that the station transmits a
load and system throughput measured over 20 s time intervals, packet in a generic (i.e., randomly chosen) slot time. This prob-
and normalized with respect to the channel rate. ability does not depend on the access mechanism (i.e., Basic
From the figure, we see that the measured throughput follows or RTS/CTS) employed. Then, by studying the events that can
closely the measured offered load for the first 260 s of sim- occur within a generic slot time, we express the throughput of
ulation, while it asymptotically drops to the value 0.68 in the both Basic and RTS/CTS access methods (as well as of a com-
second part of the simulation run. This asymptotic throughput bination of the two) as function of the computed value .
value is referred to, in this paper, as saturation throughput, and
represents the system throughput in overload conditions. Note A. Packet Transmission Probability
than, during the simulation run, the instantaneous throughput Consider a fixed number of contending stations. In satura-
temporarily increases over the saturation value (up to 0.74 in tion conditions, each station has immediately a packet available
Authorized licensed use limited to: Indian Institute of Technology Indore. Downloaded on February 06,2024 at 09:12:24 UTC from IEEE Xplore. Restrictions apply.
538 IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS, VOL. 18, NO. 3, MARCH 2000

Fig. 4. Markov Chain model for the backoff window size.

for transmission, after the completion of each successful trans- The key approximation in our model is that, at each transmis-
mission. Moreover, being all packets “consecutive,” each packet sion attempt, and regardless of the number of retransmissions
needs to wait for a random backoff time before transmitting. suffered, each packet collides with constant and independent
Let be the stochastic process representing the backoff probability . It is intuitive that this assumption results more
time counter for a given station. A discrete and integer time accurate as long as and get larger. will be referred to as
scale is adopted: and correspond to the beginning of conditional collision probability, meaning that this is the prob-
two consecutive slot times, and the backoff time counter of each ability of a collision seen by a packet being transmitted on the
station decrements at the beginning of each slot time. Note that channel.
this discrete time scale does not directly relates to the system Once independence is assumed, and is supposed to be a con-
time. In fact, as illustrated in Fig. 1, the backoff time decrement stant value, it is possible to model the bidimensional process
is stopped when the channel is sensed busy, and thus the time with the discrete-time Markov chain depicted in
interval between two consecutive slot time beginnings may be Fig. 4. In this Markov chain, the only non null one-step tran-
much longer than the slot time size , as it may include a packet sition probabilities are2
transmission. In what follows, unless ambiguity occurs, with
the term slot time we will refer to either the (constant) value ,
and the (variable) time interval between two consecutive backoff
time counter decrements.
Since the value of the backoff counter of each station depends
also on its transmission history (e.g., how many retransmis- (1)
sion the head-of-line packet has suffered), the stochastic process The first equation in (1) accounts for the fact that, at the
is non-Markovian. However, define for convenience beginning of each slot time, the backoff time is decremented.
. Let , “maximum backoff stage,” be the value such The second equation accounts for the fact that a new packet
that , and let us adopt the notation , following a successful packet transmission starts with backoff
where is called “backoff stage.” Let be the sto- 2We adopt the short notation:
chastic process representing the backoff stage of the
station at time . P fi ;k ji ;k g = P fs(t + 1) = i ; b( t + 1) = k j s(t) = i ; b( t ) = k g:
Authorized licensed use limited to: Indian Institute of Technology Indore. Downloaded on February 06,2024 at 09:12:24 UTC from IEEE Xplore. Restrictions apply.
BIANCHI: PERFORMANCE ANALYSIS OF THE IEEE 802.11 DCF 539

stage 0, and thus the backoff is initially uniformly chosen in the However, in general, depends on the conditional collision
range . The other cases model the system after an probability , which is still unknown. To find the value of
unsuccessful transmission. In particular, as considered in the it is sufficient to note that the probability that a transmitted
third equation of (1), when an unsuccessful transmission occurs packet encounters a collision, is the probability that, in a time
at backoff stage , the backoff stage increases, and the new slot, at least one of the remaining stations transmit. The
initial backoff value is uniformly chosen in the range . fundamental independence assumption given above implies that
Finally, the fourth case models the fact that once the backoff each transmission “sees” the system in the same state, i.e., in
stage reaches the value , it is not increased in subsequent steady state. At steady state, each remaining station transmits a
packet transmissions. packet with probability . This yields
Let
be the stationary distribution of the chain. We (9)
now show that it is easy to obtain a closed-form solution for this
Equations (7) and (9) represent a nonlinear system in the two
Markov chain. First, note that
unknowns and , which can be solved using numerical tech-
niques. It is easy to prove that this system has a unique solution.
In fact, inverting (9), we obtain .
(2) This is a continuous and monotone increasing function in the
range , that starts from and grows up to
Owing to the chain regularities, for each , it is . Equation defined by (7) is also continuous in
the range : continuity in correspondence of the crit-
ical value is simply proven by noting that can be
(3) alternatively written as

By means of relations (2), and making use of the fact that

, (3) rewrites as
and, therefore, . Moreover,
(4) is trivially shown to be a monotone decreasing function that
starts from and reduces up to
Thus, by relations (2) and (4), all the values are expressed . Uniqueness of the solution is now proven noting that
as functions of the value and of the conditional collision and .
probability . is finally determined by imposing the nor-
malization condition, that simplifies as follows: B. Throughput
Let be the normalized system throughput, defined as the
fraction of time the channel is used to successfully transmit pay-
load bits. To compute , let us analyze what can happen in a
randomly chosen slot time. Let be the probability that there
(5) is at least one transmission in the considered slot time. Since
stations contend on the channel, and each transmits with proba-
bility
from which
(10)
(6)
The probability that a transmission occurring on the channel
We can now express the probability that a station trans- is successful is given by the probability that exactly one station
mits in a randomly chosen slot time. As any transmission occurs transmits on the channel, conditioned on the fact that at least
when the backoff time counter is equal to zero, regardless of the one station transmits, i.e.,
backoff stage, it is
(11)

We are now able to express as the ratio

(7)
As a side note, it is interesting to highlight that, when , payload information transmitted in a slot time
(12)
i.e., no exponential backoff is considered, the probability re- length of a slot time
sults to be independent of , and (7) becomes the much simpler Being the average packet payload size, the average amount
one independently found in [9] for the constant backoff window of payload information successfully transmitted in a slot time
problem is , since a successful transmission occurs in a slot
time with probability . The average length of a slot time
(8) is readily obtained considering that, with probability ,
Authorized licensed use limited to: Indian Institute of Technology Indore. Downloaded on February 06,2024 at 09:12:24 UTC from IEEE Xplore. Restrictions apply.
540 IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS, VOL. 18, NO. 3, MARCH 2000

Fig. 5. T and T for basic access and RTS/CTS mechanisms.

the slot time is empty; with probability it contains a suc- payload size. Taking the conditional expectation on the number
cessful transmission, and with probability it con- of colliding packets, writes as follows:
tains a collision. Hence, (12) becomes

(13)

(15)
Here, is the average time the channel is sensed busy (i.e.,
the slot time lasts) because of a successful transmission, and When the probability of three or more packets simultaneously
is the average time the channel is sensed busy by each station colliding is neglected, (15) simplifies to
during a collision. is the duration of an empty slot time. Of
course, the values , and must be expressed with (16)
the same unit.
Note that the throughput expression (13) has been obtained is the period of time during which the channel is sensed
without the need to specify the access mechanism employed. busy by the noncolliding stations. We neglect the fact that the
To specifically compute the throughput for a given DCF ac- two or more colliding stations, before sensing the channel again,
cess mechanism it is now necessary only to specify the corre- need to wait an ACK Timeout, and thus the for these col-
sponding values and . liding stations is greater than that considered here (the same ap-
Let us first consider a system completely managed via the proximation holds in the following RTS/CTS case, with a CTS
basic access mechanism. Let be Timeout instead of the ACK timeout).
the packet header, and be the propagation delay. As shown Let us now consider a system in which each packet is trans-
in Fig. 5, in the basic access case we obtain mitted by means of the RTS/CTS Access mechanism. As, in
such a case, collision can occur only on RTS frames, it is (see
Fig. 5)

(14) (17)
where is the the average length of the longest packet
payload involved in a collision.
In the case all packets have the same fixed size, and the throughput expression depends on the packet size dis-
. In the general case, the payload size of each col- tribution only through its mean.
lided packet is an independent random variable . It is thus Finally, (13) can be also adopted to express the throughput of
necessary to assume a suitable probability distribution function an “Hybrid” system in which, as suggested in the standard [3],
for the packet's payload size. Let be the maximum packets are transmitted by means of the RTS/CTS mechanism
Authorized licensed use limited to: Indian Institute of Technology Indore. Downloaded on February 06,2024 at 09:12:24 UTC from IEEE Xplore. Restrictions apply.
BIANCHI: PERFORMANCE ANALYSIS OF THE IEEE 802.11 DCF 541

only if they exceed a given predetermined threshold on the Finally, noting that in the case of collision between two basic ac-
packet's payload size. More specifically, being, again, the cess packets, the probability distribution function of the length
probability distribution function of the packet size, is the of the longest packet payload involved in a collision is the square
probability that a packet is transmitted according to the basic of the conditional probability distribution function of the packet
access mechanism (i.e., the packet size is lower than ), while size distribution
is the probability that a packet is transmitted via the
RTS/CTS mechanism. For convenience, let us indicate with
(23)
(18)
By substituting (21), (22), and (23) in (20), we finally obtain
the RTS/CTS overhead for a successful packet transmission. It is
easy to recognize that, for the described hybrid access scheme,
it is

(19) (24)

To compute in the case of the Hybrid Access For simplicity, in the rest of this paper we restrict our numer-
scheme, we rely on the simplifying assumption that the proba- ical investigation to the case of fixed packet size, and therefore
bility of a collision of more than two packets in the same slot we will evaluate the performance of systems in which all sta-
time is negligible. Hence, three possible collision cases may tions operate either according to the basic access mechanism or
occur: 1) collision between two RTS frames, with probability according to the RTS/CTS mechanism (i.e., never operating in
; 2) collision between two packets transmitted via the hybrid mode.)3
basic access, with probability , and 3) collision between
a basic access packet and an RTS frame. Hence, indicating with V. MODEL VALIDATION
and the respective average collision To validate the model, we have compared its results with that
durations, we obtain obtained with the 802.11 DCF simulator used in [9]. Ours is
an event-driven custom simulation program, written in the C++
programming language, that closely follows all the 802.11 pro-
tocol details for each independently transmitting station. In par-
(20) ticular, the simulation program attempts to emulate as closely
as possible the real operation of each station, including propa-
gation times, turnaround times, etc.
The average collision durations adopted in (20) detail as fol- The values of the parameters used to obtain numerical results,
lows. Let be the extra for both the analytical model and the simulation runs, are sum-
length of the packet header with respect of the RTS frame, and marized in Table II. The system values are those specified for the
let . The value has been already frequency hopping spread spectrum (FHSS) PHY layer [3]. The
computed in the case of (17), and can be rewritten with channel bit rate has been assumed equal to 1 Mbit/s. The frame
new notation as sizes are those defined by the 802.11 MAC specifications, and
the PHY header is that defined for the FHSS PHY. The values
(21) of the ACK_Timeout and CTS_Timeout reported in Table II,
and used in the simulation runs only (our analysis neglects the
To compute the average length of a collision between an RTS effect of these timeouts) are not specified in the standard, and
frame and a basic access packet, let us note that, according to they have been set equal to 300 µs. This numerical value has
the numerical values provided by the standard [3], the length been chosen as it is sufficiently long to contain a SIFS, the ACK
of an RTS frame is always lower than the packet header size, transmission and a round trip delay.
or, in other words, the value defined above is strictly Unless otherwise specified, we have used in the simulation
positive. Thus, the average length of such a collision is given runs a constant packet payload size of 8184 bits, which is about
by the average amount of time the channel is kept busy by the one fourth of the maximum MPDU size specified for the FHSS
unsuccessful transmission of the basic access packet. Since PHY, while it is the maximum MPDU size for the DSSS PHY.
is the conditional probability distri- Fig. 6 shows that the analytical model is extremely accurate:
bution function of the payload size of the packets transmitted analytical results (lines) practically coincide with the simulation
according to the basic access mechanism, we readily obtain 3A detailed performance analysis of the hybrid mode requires to assume one
or more suitable probability distribution functions for the packet's payload size,
and also to determine the sensitivity of the throughput on the assumed distri-
(22) butions. Such a straightforward, but lengthy, study is out of the scopes of the
present work.
Authorized licensed use limited to: Indian Institute of Technology Indore. Downloaded on February 06,2024 at 09:12:24 UTC from IEEE Xplore. Restrictions apply.
542 IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS, VOL. 18, NO. 3, MARCH 2000

results (symbols), in both basic access and RTS/CTS cases. All TABLE II
simulation results in the plot are obtained with a 95% confidence FHSS SYSTEM PARAMETERS AND ADDITIONAL PARAMETERS USED TO
OBTAIN NUMERICAL RESULTS
interval lower than 0.002. Negligible differences, well below
1%, are noted only for a small number of stations (results for
the extreme case of as low as 2 and 3 stations are tabulated in
Table III).

VI. MAXIMUM SATURATION THROUGHPUT

The analytical model given above is very convenient to de-
termine the maximum achievable saturation throughput. Let us
rearrange (13) to obtain

(25)

As , and , are constants, the throughput is max-

imized when the following quantity is maximized:

(26)

where is the duration of a collision measured in slot

time units . Taking the derivative of (26) with respect to , and
imposing it equal to 0, we obtain, after some simplifications, the
following equation:

(27)

Under the condition

Fig. 6. Saturation Throughput: analysis versus simulation.

TABLE III
ANALYSIS VERSUS SIMULATION: COMPARISON FOR A VERY LOW NUMBER OF
STATIONS—W = 32; m = 3
holds, and yields the following approximate solution:

(28)

Equation (27) and its approximate solution (28) are of funda-

mental theoretical importance. In fact, they allow to explicitly is given by (8), and therefore the backoff window that maxi-
compute the optimal transmission probability that each sta- mizes the system throughput is readily found as
tion should adopt in order to achieve maximum throughput per-
formance within a considered network scenario (i.e., number of
stations ). In other words, they show that (within a PHY and
an access mechanism, which determine the constant value ) Refer to [9] for an extensive discussion related to the problem
maximum performance can be, in principle, achieved for every of estimating the value .
network scenario, through a suitable sizing of the transmission Unfortunately, in the 802.11 standard, the values and
probability in relation to the network size. are hardwired in the PHY layer details (see Table I for the stan-
However, (7) and (9) show that depends only on the network dardized values), and thus they cannot be made dependent on
size and on the system parameters and . As is not a . As a consequence of this lack of flexibility, the throughput
directly controllable variable, the only way to achieve optimal in some network scenarios can be significantly lower than the
performance is to employ adaptive techniques to tune the values maximum achievable.
and (and consequently ) on the basis of the estimated Figs. 7 and 8 show the maximum throughput theoretically
value of . achievable by the DCF protocol in both the cases of basic access
This problem has been specifically considered in [9] for the and RTS/CTS mechanisms. The values reported in these figures
case of fixed backoff window size (i.e., ). In such a case, have been obtained assuming the system parameters reported in
Authorized licensed use limited to: Indian Institute of Technology Indore. Downloaded on February 06,2024 at 09:12:24 UTC from IEEE Xplore. Restrictions apply.
BIANCHI: PERFORMANCE ANALYSIS OF THE IEEE 802.11 DCF 543

Fig. 8. Throughput versus the transmission probability  for the RTS/CTS

Fig. 7. Throughput versus the transmission probability  for the basic access mechanism.
method.

TABLE IV

=1
Table II. The figure reports also the different throughput values COMPARISON BETWEEN MAXIMUM THROUGHPUT AND THROGHPUT
obtained in the case of exact and approximate solution for . As RESULTING FROM APPROXIMATE SOLUTION (28)—THE CASE n
IS OBTAINED FROM (31)
the maximum is very smooth, even a nonnegligible difference in
the estimate of the optimal value leads to similar throughput
values. The accuracy of the throughput obtained by the approxi-
mate solution is better testified by the numerical values reported
in Table IV. Note that the agreement is greater in the basic ac-
cess case, as is greater.
A surprising result is that the maximum throughput achiev-
able by the basic access mechanism is very close to that
achievable by the RTS/CTS mechanism. Moreover, the max-
imum throughput is practically independent of the number
of stations in the wireless network. This is easily justified by
noting that the throughput formula can be approximated as
follows. Let , and let us use the approximate
solution . For sufficiently large
TABLE V
VALUES T AND T
MEASURED IN BITS AND IN 50 s SLOT TIME UNITS,
FOR THE CONSIDERED SYSTEM PARAMETERS, FOR BOTH BASIC AND
(29) RTS/CTS ACCESS METHODS

(30)
The maximum achievable throughput can thus be approx-
imated as
decrease in the throughput for the basic access case than for
the RTS/CTS case. Hence, we expect (see quantitative results
(31) in the following Section VII) a much lower dependence of the
RTS/CTS throughput on the system engineering parameters
with respect of the basic access throughput.
which results to be independent of . Using the numerical values
of Table V, we obtain for the basic access mecha-
nism, and for the RTS/CTS mechanism. The re- VII. PERFORMANCE EVALUATION
sulting maximum throughput approximation values are reported Unless otherwise specified, the following results have been
in Table IV under the label . obtained assuming the parameters reported in Table II and, in
An advantage of the RTS/CTS scheme is that the throughput particular, assuming a constant payload size bits.
is less sensitive on the transmission probability . In fact, we Fig. 6 shows that the throughput for the basic access scheme
see from Figs. 7 and 8 (note the different axis scale) that strongly depends on the number of stations in the network. In
a small variation in the optimal value of leads to a greater particular, the figure shows that, in most cases, the greater is the
Authorized licensed use limited to: Indian Institute of Technology Indore. Downloaded on February 06,2024 at 09:12:24 UTC from IEEE Xplore. Restrictions apply.
544 IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS, VOL. 18, NO. 3, MARCH 2000

Fig. 9. Saturation Throughput versus initial contention window size for the
basic access mechanism. Fig. 10. Saturation Throughput versus initial contention window size for the
RTS/CTS mechanism.

network size, the lower is the throughput. The only partial ex-
ception is the case . For such an initial contention Of more practical interest is the case of small values of ,
window size, the throughput is comparable in networks with and particularly in correspondence of the values ,
five to ten stations, although it smoothly decreases as the net- and (i.e., those standardized for the three PHY—see Table I).
work size increases. The same figure shows that performance Figs. 9 and 10 show that the two access mechanisms achieve
impairment does not occur for the RTS/CTS mechanism when a significantly different operation. In the case of the basic ac-
increases. In fact, the throughput is practically constant for cess mechanism, reported in Fig. 9, the system throughput in-
, and even increases with the number of mobile sta- creases as long as gets closer to 64. Moreover, the throughput
tions when . significantly decreases as the number of stations increases. On
To investigate the dependency of the throughput from the ini- the contrary, Fig. 10 shows that the throughput obtained with
tial contention window size we have reported in Figs. 9 and the RTS/CTS mechanism is almost independent of the value
10 the saturation throughput versus the value for, respec- , and, in this range, it is furthermore almost insensi-
tively, the basic access and the RTS/CTS mechanisms. In both tive on the network size.
figures, we have assumed a number of backoff stages equal to 6, This surprising independence is quantitatively explained as
i.e., . The figures report four different network follows. Dividing numerator and denominator of (13) by ,
sizes, i.e., number of stations equal to 5, 10, 20, and 50. we obtain
Fig. 9 shows that the throughput of the basic access mecha-
nism highly depends on , and the optimal value of depends (33)
on the number of terminals in the network. For example, an high
value of (e.g., 1024) gives excellent throughput performance
in the case of 50 contending stations, while it drastically penal- The denominator of (33) expresses the average amount of time
izes the throughput in the case of small number (e.g., 5) of con- spent on the channel in order to observe the successful transmis-
tending stations. This behavior is seen also in Fig. 10, where the sion of a packet payload. This time is further decomposed into
RTS/CTS mechanism is employed. Large values of may, in three components.
fact, limit the throughput of a single station, which, when alone is the time spent in order to successfully transmit a packet.
in the channel is bounded by Table V reports the numerical values for and , com-
puted according to (14) and (17), in the assumption of system
and channel parameters of Table II. The difference between
(32) and (586 bits) is the additional overhead introduced by the
RTS/CTS mechanism.
The second term at the denominator of (33) does not depend
where and are the average packet payload and the av- on the access mechanism employed, and represents the amount
erage channel holding time in case of successful transmission. of time the channel is idle, per successful packet transmission.
Equation (32) is directly obtained from (13) of Section IV-B by In fact, is the average number of slot times spent on
observing that, as there are no other stations which can collide the channel in order to have a successful transmission. Of those
with the considered one, the probability of success is equal to slot times, a fraction is empty, and each empty slot time
1. In addition, the probability that a transmission occurs on lasts . The average number of idle slot times per packet trans-
the channel is equal to the probability that the station trans- mission, i.e., , is plotted in Fig. 11 versus the
mits. Being the conditional collision probability equal to 0, network size, for three different values of the initial contention
is given by (8). window . We see that, for and , the amount
Authorized licensed use limited to: Indian Institute of Technology Indore. Downloaded on February 06,2024 at 09:12:24 UTC from IEEE Xplore. Restrictions apply.
BIANCHI: PERFORMANCE ANALYSIS OF THE IEEE 802.11 DCF 545

of idle slot times per packet transmission is very low, particu-

larly when compared with the values given in Table V. This
value becomes significant only when gets greater (the case
is reported in the figure) and the number of stations
in the network is small.
Finally, the third term at the denominator of (33) repre-
sents the time wasted on the channel because of collisions,
per successful packet transmission. In fact, is the
average number of collided transmissions per each successful
transmission, which is multiplied by , i.e., the amount of
time the channel is held by a collision. Table V shows that
the the RTS/CTS mechanism significantly reduces the time
spent during a collision, with respect to the basic access mech-
anism. This reduction is extremely effective when the system
parameter and the network size lead to a large collision
probability. This fact is graphically shown in Fig. 12. This
figure reports the average amount of time spent in collisions,
per successful packet transmission, normalized with respect Fig. 11. Average number of idle slot times per successful packet transmission.
to the value . It shows that, for the basic access mechanism,
the amount of channel time wasted in collisions is extremely
large for a small value and a large number of stations in the
network. Conversely, the additional amount of time wasted in
collisions is negligible for the RTS/CTS mechanism, regardless
of the values and . This explains the surprising constant
RTS/CTS throughput in any practical system and network
operation conditions.
Fig. 13 shows that the dependence of the throughput from the
maximum number of backoff stages is marginal. The figure
reports the cases of both Basic and RTS/CTS access schemes,
with (similar behaviour is observed for other values of
the parameter ) and . The points in the box indicate
the throughput achieved when , i.e., in correspondence
of the standardized engineering parameters of the DSSS PHY
(Table I). We see that the choice of does not practically affect
the system throughput, as long as is greater than four or five.
The only case in which the throughput still grows, for rela- Fig. 12. Average number of slot time units wasted on the channel because of
tively large, is the basic access mechanism with a large network packet collision, per successful packet transmission.
size.
Our model allows to obtain other measures of interest. The Columns in boldface type correspond to the standardized slot
conditional collision probability is the probability, seen by the time length for the related PHY. The marginal dependence of
station, that its transmitted packet collides. Owing to the model's the throughput on the slot time size is related to the fact, com-
key assumption of independence at each retransmission, the av- mented above by means of (33) and Fig. 11, that the number
erage number of transmissions that each station must perform of idle slot times per packet transmission is extremely small. A
in order to successfully complete a packet transmission is given change of has the only effect to multiply by a constant value
by . This value is reported in Fig. 14, obtained with the the amount of idle channel time per packet transmission. How-
same system parameters of Figs. 9 and 10. Fig. 14 shows that ever, for any practical value of and , the amount of idle
the number of transmissions per packet significantly increases channel time remains marginal with respect to the time spent in
as the initial backoff window reduces, and as the network transmission and collision. This result is of fundamental impor-
size increases. tance for the future development of higher bit rate physical layer
At a first glance, it might seem that the throughput perfor- recommendations, as the slot time size is difficultly scalable.
mance of the 802.11 protocol strongly depends on the slot time Finally, let us add some considerations regarding the depen-
size . In particular, the lower is , the better is the expected per- dence of the access method on the packet length. It is often
formance. Instead, we note that, as far as saturation throughput qualitatively stated that the RTS/CTS mechanism is effective
performance is concerned, its dependence on the slot time size when the packet size increases. This is justified in Fig. 15. This
is only marginal. Table VI reports results for three different figure reports the system throughput for both basic access and
system configurations corresponding to the different PHY’s. RTS/CTS cases, for two different network sizes ( and
Results are obtained for both basic access and RTS/CTS mecha- ), and for three different configuration parameters, re-
nism, and for two different network sizes of 10 and 50 stations. ferred to as FH, DS and IR, corresponding to the three PHY's
Authorized licensed use limited to: Indian Institute of Technology Indore. Downloaded on February 06,2024 at 09:12:24 UTC from IEEE Xplore. Restrictions apply.
546 IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS, VOL. 18, NO. 3, MARCH 2000

TABLE VI
DEPENDENCE OF THE SATURATION THROUGHPUT ON THE SLOT TIME

Fig. 13. Throughput versus the maximum number of backoff stages: W = 32.

Fig. 15. Throughput versus packet size for the standardized configuration
parameters.

Fig. 14. Average number of transmissions per packet.

Let now the overhead introduced by the
RTS/CTS mechanism, and let be the extra
reference values and Slot time size reported length of the packet header with respect of the RTS frame size
in Table I. It is no more a surprise that the RTS/CTS mechanism (according to the values of Table II, bits, and
achieves very similar performance in all the considered cases. bits). Indicating the packet payload with the variable ,
This is due to the fact that the throughput performance margin- condition (34) yields
ally depends on the slot time, as shown in Table VI, and on the
fact that the RTS/CST scheme is negligibly dependent on the
network size and on the minimum contention window size. (35)
In the assumption of fixed packet payload size, it is very easy
to quantify the threshold value for the packet size over which it
is convenient to switch to the RTS/CTS mechanism. In fact, let The threshold value over which it is convenient to switch
us indicate with and the throughput achieved respec- to the RTS/CTS scheme is plotted versus the network size in
tively by the basic access and RTS/CTS mechanism in the same Fig. 16, for the three possible sets of parameters specified for the
system parameters and network size conditions. From (33), the different PHY’s. This figure shows that the threshold is highly
inequality dependent on the PHY employed. This is not a consequence of
the different slot time size , which does not affect (35). In-
stead, it is a direct consequence of the different initial contention
window sizes adopted (see Table I). The lower the value ,
implies that the greater is the performance impairment of the basic access
scheme (see Fig. 9), and the greater (and thus for more packet
size cases, as shown in Fig. 15) is the advantage of the RTS/CTS
(34)
scheme.
Authorized licensed use limited to: Indian Institute of Technology Indore. Downloaded on February 06,2024 at 09:12:24 UTC from IEEE Xplore. Restrictions apply.
BIANCHI: PERFORMANCE ANALYSIS OF THE IEEE 802.11 DCF 547

sizes. When the capability of the RTS/CTS scheme to cope with

hidden terminals is accounted, we conclude that this access
method should be used in the majority of the practical cases.

ACKNOWLEDGMENT
The author wishes to thank the anonymous refereers for their
helpful comments that have significantly improved the quality
of the presentation.

REFERENCES
[1] K. Pahlavan and A. H. Levesque, “Wireless data communications,”
Proc. IEEE, vol. 82, pp. 1398–1430, Sept. 1994.
[2] A. De Simone and S. Nanda, “Wireless data: Systems, standards, ser-
vices,” J. Wireless Networks, vol. 1, no. 3, pp. 241–254, Feb. 1996.
[3] IEEE Standard for Wireless LAN Medium Access Control (MAC) and
Physical Layer (PHY) Specifications, Nov. 1997. P802.11.
[4] L. Kleinrock and F. Tobagi, “Packet switching in radio channels, Part
Fig. 16. Packet payload threshold over which the RTS/CTS mechanism is II—The hidden terminal problem in carrier sense multiple access and
advantageous. the busy tone solution,” IEEE Trans. Commun., vol. COM-23, no. 12,
pp. 1417–1433, Dec. 1975.
[5] H. S. Chhaya and S. Gupta, “Performance modeling of asynchronous
Moreover, Fig. 16 runs counter to the “known” fact that the data transfer methods of IEEE 802.11 MAC protocol,” Wireless Net-
works, vol. 3, pp. 217–234, 1997.
RTS/CTS mechanism should be employed when the packet size [6] K. C. Huang and K. C. Chen, “Interference analysis of nonpersistent
exceeds a given (meaning fixed) threshold. Instead, it shows that CSMA with hidden terminals in multicell wireless data networks,” in
such a threshold strongly depends on the network size, and par- Proc. IEEE PIMRC, Toronto, Canada, Sept. 1995, pp. 907–911.
[7] B. P. Crow, “Performance evaluation of the IEEE 802.11 Wireless Local
ticularly it significantly decreases when the number of stations Area Network Protocol,” M.S. thesis, Dept. Electrical and Computer
in the network increases. For example, in the case of 50 stations, Eng., Univ. Arizona, Tucson, AZ, 1996.
the threshold is equal to about 1470 bits for the infrared PHY, [8] J. Weinmiller, M. Schlager, A. Festag, and A. Wolisz, “Performance
study of access control in wireless LANs IEEE 802.11 DFWMAC and
while it is as low as 820 bits for the frequency hopping PHY. ETSI RES 10 HIPERLAN,” Mobile Networks and Applicat., vol. 2, pp.
The same threshold raises, respectively, to about 10065 bits and 55–67, 1997.
3160 bits when the network is composed by five stations only. [9] T. S. Ho and K. C. Chen, “Performance evaluation and enhancement
of the CSMA/CA MAC protocol for 802.11 wireless LAN’s,” in Proc.
IEEE PIMRC, Taipei, Taiwan, Oct. 1996, pp. 392–396.
VIII. CONCLUSION [10] F. Cali, M. Conti, and E. Gregori, “IEEE 802.11 wireless LAN: Capacity
analysis and protocol enhancement,” presented at the INFOCOM'98,
In this paper, we have presented a simple analytical model to San Francisco, CA, Mar. 1998.
[11] G. Bianchi, L. Fratta, and M. Oliveri, “Performance analysys of IEEE
compute the saturation throughput performance of the 802.11 802.11 CSMA/CA medium access control protocol,” in Proc. IEEE
Distributed Coordination Function. Our model assumes a finite PIMRC, Taipei, Taiwan, Oct. 1996, pp. 407–411.
number of terminals and ideal channel conditions. The model [12] G. Bianchi, “IEEE 802.11—Saturation throughput analysis,” IEEE
Commun. Lett., vol. 2, pp. 318–320, Dec. 1998.
is suited for any access scheme employed, i.e., for both basic [13] D. Bertsekas and R. Gallager, Data Networks. Englewood Cliffs, NJ:
access and RTS/CTS Access mechanisms, as well as for a com- Prentice-Hall, 1987.
bination of the two. Comparison with simulation results shows
that the model is extremely accurate in predicting the system
throughput. Giuseppe Bianchi received the “Laurea” degree in
Using the proposed model, we have evaluated the 802.11 electronic engineering from Politecnico di Milano,
throughput performance. We have shown that performance of Milano, Italy, in 1990, and the specialization degree
in information technology from CEFRIEL, Milano,
the basic access method strongly depends on the system pa- in 1991.
rameters, mainly minimum contention window and number of He was a Researcher at CEFRIEL from 1991 to
stations in the wireless network. Conversely, performance is 1993, and Assistant Professor at the Politecnico di
Milano from 1994 to 1998. He spent 1992 as a Visitor
only marginally dependent on the system parameters when the Researcher at Washington University, St. Louis, MO,
RTS/CTS mechanism is considered. and spent 1997 as a Visiting Professor at Columbia
The RTS/CTS mechanism has proven its superiority in most University, New York. Since November 1998, he has
been Associate Professor at the University of Palermo, Italy. His research in-
of the cases. Notable is the advantage of the RTS/CTS scheme terests include design and performance evaluation of broadband and wireless
in large network scenarios, even with fairly limited packet networks and systems.

Authorized licensed use limited to: Indian Institute of Technology Indore. Downloaded on February 06,2024 at 09:12:24 UTC from IEEE Xplore. Restrictions apply.