IEEE 802.

11 Protocol architecture and

Management
• The IEEE 802.11 standard only covers the physical layer PHY and
medium access layer MAC
• The physical layer is subdivided into
- Physical layer convergence protocol (PLCP) and
- Physical medium dependent sublayer (PMD)

• PLCP sublayer provides

- A carrier sense signal, clear channel assessment
(CCA)
-PHY Service Access Point (SAP) independent of the
transmission technology
• PMD sublayer handles modulation and encoding/decoding of
signals
• The standard also specifies management layers and the station
management
• MAC layer comprise
-Medium access
-Fragmentation of user data and
-Encryption
• The MAC management supports
- Association and re-association of a station to an access point
and
- Roaming between different access points
- Controls authentication mechanisms, encryption,
synchronization of a station with regard to an access point
- Power management to save battery power
• MAC management also maintains the MAC management information
base (MIB)
• The main tasks of the PHY management include
- Channel tuning and
- PHY MIB maintenance
• Station management interacts with both management layers and is
responsible for additional higher layer functions (e.g., control of bridging
and interaction with the distribution system in the case of an access point)

Physical layer:
• IEEE 802.11 supports three different physical layers:
- One layer based on infra red and
- Two layers based on radio transmission
• All PHY variants include the provision of the Clear Channel
Assessment signal (CCA)
• This is needed for the MAC mechanisms controlling medium access
and indicates if the medium is currently idle
• The PHY layer offers a service access point (SAP) with 1 or 2 Mbit/s
transfer rate to the MAC layer
 Infra red
• The PHY layer, which is based on infra red (IR) transmission, uses near
visible light at 850–950 nm
• The standard does not require a line-of-sight between sender and receiver,
but should also work with diffuse light
• This allows for point-to-multipoint communication
• The maximum range is about 10 m if no sunlight or heat sources interfere
with the transmission
• Typically, such a network will only work in buildings, e.g., classrooms,
meeting rooms etc
• Frequency reuse is very simple – a wall is more than enough to shield one
IR based IEEE 802.11 network from another
• Today, no products are available that offer infra red communication based
on 802.11
• Proprietary products offer, e.g., up to 4 Mbit/s using diffuse infra red light
• Alternatively, directed infra red communication based on IrDA can be used
(IrDA, 2002).
IEEE 802.11 MEDIUM ACCESS CONTROL LAYER
• MAC layer has to control medium access and offer support for
roaming, authentication and power conservation
• Basic services provided by the MAC layer
- Asynchronous data service (mandatory) and
- Time-bounded service (optional)
• Asynchronous service is offered in ad-hoc network mode
• Both service types can be offered using an infrastructure-based
network along with access point coordinating medium access
• Asynchronous service supports broadcast and multi-cast packets,
and packet exchange is based on a ‘best effort’ model (no delay
bounds) can be given for transmission
• Three basic access mechanisms have been defined for IEEE 802.11
- Basic DFWMAC-DCF using CSMA/CA
- DFWMAC-DCF with RTS/CTS extension
- DFWMAC-PCF with polling
• First two methods are Distributed coordination function (DCF),
the third method is called Point coordination function (PCF)
• DCF only offers asynchronous service, while PCF offers both
asynchronous and time-bounded service but needs an access point to
control medium access and to avoid contention
• The MAC mechanisms are also called Distributed foundation
wireless medium access control (DFWMAC)
• For all access methods, several parameters for controlling the
waiting time before medium access are important
- SIFS
- PIFS
- DIFS
• Values of the parameters depend on the PHY and are defined in
relation to a Slot time
• Slot time is derived from the medium propagation delay, transmitter
delay and other PHY dependent parameters
• Slot time is 50 μs for FHSS and 20 μs for DSSS
Medium access and inter-frame spacing
• The medium can be busy or idle (which is detected by the CCA)
• If the medium is busy this can be due to data frames or other
control frames
• During a contention phase several nodes try to access the medium
Short inter-frame spacing (SIFS):
• The shortest waiting time for medium access (so the highest
priority) is defined for short control messages, such as
acknowledgements of data packets or polling responses
• For DSSS SIFS is 10 μs and for FHSS it is 28 μs

PCF inter-frame spacing (PIFS):

• A waiting time between DIFS and SIFS (and thus a medium
priority) is used for a time-bounded service
• An access point polling other nodes only has to wait PIFS for
medium access
• PIFS is defined as SIFS plus one slot time
DCF inter-frame spacing (DIFS):
• This parameter denotes the longest waiting time and has the lowest
priority for medium access
• This waiting time is used for asynchronous data service within a
contention period
• DIFS is defined as SIFS plus two slot times
Consider in IEEE 802.11, SIFS is 2 ms. Calculate the PIFS and
DIFS. 1 Time slot = 1 ms

PIFS = SIFS + 1 Time slot = 3ms

DIFS = SIFS + 2 Time slots = 4ms
 Basic DFWMAC-DCF using CSMA/CA
• Access mechanism of IEEE 802.11 is based on carrier sense
multiple access with collision avoidance (CSMA/CA)
• It is a random access scheme with carrier sense and collision
avoidance through random backoff
• If the medium is idle for atleast the duration of DIFS, a node can
access the medium at once
• This allows for short access delay under light load
• But as more and more nodes try to access the medium, additional
mechanisms are needed
• If the medium is busy, nodes have to wait for the duration of DIFS,
then enters into a contention phase
• Each node now chooses a random backoff time within a
contention window and delays medium access for the random
amount of time
• The node continues to sense the medium
• As soon as a node senses the channel is busy, it has lost this cycle and
has to wait for the next chance (until the medium is idle again for at
least DIFS)
• If the randomized additional waiting time for a node is over and the
medium is still idle, the node can access the medium immediately (no
other node has a shorter waiting time)
• Additional randomly distributed delay helps to avoid collisions –
otherwise all stations would try to transmit data after waiting for
the medium becoming idle again plus DIFS
• Independent of the overall time a node has already waited for
transmission; each node has the same chances for transmitting data
in the next cycle (CSMA/CA mechanism is not fair)
• To provide fairness, IEEE 802.11 adds a backoff timer
• Each node selects a random waiting time within the range of the
contention window
• If a certain station does not get access to the medium in the first cycle,
it stops its backoff timer, waits for the channel to be idle again for
DIFS and starts the counter again
• As soon as the counter expires, the node accesses the medium
• This means that deferred stations do not choose a randomized
backoff time again, but continue to count down
• Stations that have waited longer have the advantage over
stations that have just entered, in that they only have to wait for
the remainder of their backoff timer from the previous cycle(s)
Basic DFWMAC–DCF with several competing senders
•Station3 has the first request from a higher
layer to send a packet
• Station senses the medium, waits for DIFS
and accesses the medium
• Station1, station2, and station5 have to wait
at least until the medium is idle for DIFS
again after station3 has stopped sending
• All three stations choose a backoff time
within the contention window and start
counting down their backoff timers
• Random backoff time of station1 and 5 is
sum of boe (the elapsed backoff time) and
bor
• Station2 has a total backoff time of boe only
and gets access to the
medium first
• No residual backoff time for station2
• The backoff timers of station1 and station5
stop and the stations store their residual
backoff times
•A new station has to choose its backoff time
from the whole contention window, the two
old stations have statistically smaller backoff
values
• The older values are on average lower than the new ones
• Station4 also wants to send a packet, after DIFS waiting time
• Two stations accidentally have the same backoff time, no matter whether
remaining or newly chosen (Results in a collision)
• Station1 stores its residual backoff time again
• A collision triggers a retransmission with a new random selection of the
backoff time
• Retransmissions are not privileged

Disadvantages:
• Access scheme has problems under heavy or light load
• Depending on the size of the contention window (CW), the random values
can either be too close together (causing too many collisions) or the values
are too high (causing unnecessary delay)
• Small CW causes a higher load on the medium (collision)
• There is a possibility to choose the same random backoff time using a
large CW
• Under a light load, a small CW ensures shorter access delays
• This algorithm is also called exponential backoff
PROBLEM
802.11 - CSMA/CA access Unicast Data Transfer
802.11 - CSMA/CA access Unicast Data Transfer
• A sender is accessing the medium and sending its data.
• The receiver answers directly with an acknowledgement (ACK)
• Receiver accesses the medium after waiting for a duration of SIFS so no
other station can access the medium in the meantime and cause a collision
• The other stations have to wait for DIFS plus their backoff time
• Acknowledgement ensures the correct reception (correct checksum CRC
at the receiver) of a frame on the MAC layer
• It may be important in error-prone environments (wireless environment)
• If no ACK is returned, the sender automatically retransmits the frame
• But now the sender has to wait again and compete for the access right
• There are no special rules for retransmissions
• The number of retransmissions is limited and final failure is reported to the
higher layer
 HIDDEN TERMINALS

Hidden terminals
• A sends to B, C cannot receive A
• C wants to send to B, C senses a “free” medium (CS fails)
collision at B, A cannot receive the collision (CD fails)
A is “hidden” for C
• Hidden terminals may cause collisions
 EXPOSED TERMINALS

•B sends to A, C wants to send to another terminal (not A or

B)
•C has to wait, CS signals a medium in use but A is outside
the radio range of C, therefore waiting is not necessary, C is
“exposed” to B
• It causes unnecessary delay
Hidden Terminal Problem
• This CTS is now heard by C and the medium for future use by
A is now reserved for the duration of the transmission.
• After receiving a CTS, C is not allowed to send anything for
the duration indicated in the CTS toward B.
• A collision cannot occur at B during data transmission and the
hidden terminal problem is solved – provided that the
transmission conditions remain the same.
• Another station could move into the transmission range of B
after the transmission of CTS
• Still, collisions can occur during the sending of an RTS.
• Both A and C could send an RTS that collides at B.
• RTS is very small compared to the data transmission, so the
probability of a collision is much lower.
• B resolves this contention and acknowledges only one station
in the CTS
Can MACA also help to solve the
exposed terminal problem?
• With MACA, B has to transmit an RTS first containing the name of
the receiver (A) and the sender (B).
• C does not react to this message as it is not the receiver, but A
acknowledges using a CTS which identifies B as the sender and A as
the receiver of the following data transmission.
• C does not receive this CTS and concludes that A is outside the
detection range.
• C can start its transmission assuming it will not cause a collision at
A.
• The problem with exposed terminals is solved without fixed access
patterns or a base station.
Disadvantages:
• One problem of MACA is clearly the overheads associated with the
RTS and CTS transmissions – for short and time-critical data
packets, this is not negligible.
• MACA also assumes symmetrical transmission and reception
conditions.
• Otherwise, a strong sender, directed antennas etc. could counteract
the above scheme.
 DFWMAC-DCF with RTS/CTS extension
• Hidden terminal problem occurs if one station can receive two others, but
those stations cannot receive each other (collision)
• It is resolved by using RTS/CTS
• Net allocation Vector (NAV) is used in this scheme

Sending unicast packets

– station can send RTS with reservation parameter after waiting for DIFS
(reservation determines amount of time the data packet needs the
medium)
– acknowledgement via CTS after SIFS by receiver (if ready to receive)
– sender can now send data at once, acknowledgement via ACK
– other stations store medium reservations distributed via RTS and CTS
• Hidden Node provisions for Contention free access
 DFWMAC-DCF with RTS/CTS extension
•After waiting for DIFS, the sender can
issue a Request to Send (RTS) control
packet
•RTS packet is not given any higher
priority compared to other data packets
•The RTS packet includes the receiver
of the data transmission and the
duration of the whole data transmission
(Data+ACK)
• Every node receiving this RTS set its
net allocation vector (NAV) in
accordance with the duration field
•NAV specifies the earliest point at
which the station can try to access the
medium again
•If the receiver of the data transmission
receives the RTS, it answers with a
clear to send (CTS) message after
waiting for SIFS
• CTS contains the duration field and
all stations receiving this packet from
the receiver of the intended data
transmission have to adjust their NAV
• If RTS packet is transmitted, then the latter set of receivers need not be the
same as the first set
• All the nodes within the receiving distance around the senders and
receivers have to wait more time before accessing the medium
• This mechanism reserves the medium for one sender exclusively (Virtual
Reservation Scheme)
Inference:
• Sender can send the data after SIFS
• The receiver waits for SIFS after receiving the data packet and then
acknowledges whether the transfer was correct
• The transmission has now been completed, the NAV in each node marks
the medium as free and the standard cycle can start again
Disadvantages:
• Collisions can only occur at the beginning while the RTS is sent
• Waste of bandwidth and higher delay
PROBLEM
Wireless LANs have bit error rates in transmission
• If the frame length is same, the probability of an erroneous frame is much
higher for wireless links
• One way to decrease the error probability of frames is to use shorter frames
(the bit error rate is same but now only short frames are destroyed and the
frame error rate decreases)
• MAC layer should have the possibility of adjusting the transmission
frame size to the current error rate on the medium

IEEE 802.11 standard specifies a fragmentation mode :

• A sender can send an RTS control packet to reserve the medium after a
waiting time of DIFS
• RTS packet includes the duration for the transmission of the first
fragment and the corresponding acknowledgement
• A certain set of nodes may receive this RTS and set their NAV according
to the duration field
• The receiver answers with a CTS, including the duration of the
transmission up to the acknowledgement
• A (possibly different) set of receivers gets this CTS message and sets the
NAV
 Fragmentation of user data
• The sender can now send the first
data frame, frag1,after waiting only
for SIFS
• Fragmentation mode includes another
duration value in the frame frag1
• This duration field reserves the
medium for the duration of the
second fragment and its
acknowledgement
• Several nodes may receive this
reservation and adjust their NAV
•If all nodes are static and
transmission conditions have not
changed, then the set of nodes
receiving the duration field in frag1
should be the same as the set that has
received the initial reservation in the
RTS control packet
•Due to the mobility of nodes and
changes in the environment, this could
also be a different set of nodes
• The receiver of frag1 answers
directly after SIFS with the
acknowledgement packet ACK1
including the reservation for the next
transmission
PROBLEM
 DFWMAC-PCF with polling
• The above two methods cannot guarantee a maximum access delay
or minimum transmission bandwidth
• To provide a time-bounded service, the standard specifies a point
coordination function (PCF) on top of the standard DCF
mechanisms
• Using PCF requires an access point that controls medium access and
polls the single nodes
• The point co-ordinator in the access point splits the access time into
super frame periods
• Super frame comprises a contention free period and a contention
period
• The contention period can be used for the two access mechanisms
(Basic DFWMAC-DCF using CSMA/CA and DFWMAC-DCF with
RTS/CTS extension)
DFWMAC-PCF with polling
• At time t0 the contention-free period of the super frame should start, but
another station is still transmitting data (i.e., the medium is busy)
• After the medium has been idle until t1, the point coordinator has to wait
for PIFS before accessing the medium
• As PIFS is smaller than DIFS, no other station can start sending earlier
• The point coordinator now sends data D1 downstream to the first wireless
station
• This station can answer at once after SIFS
• After waiting for SIFS again, the point coordinator can poll the second
station by sending D2
• This station may answer upstream to the coordinator with data U2
• Polling continues with the third node
• This time the node has nothing to answer and the point coordinator will not
receive a packet after SIFS
• After waiting for PIFS, the coordinator can resume polling the stations
• Finally, the point coordinator can issue an end marker (CFend), indicating
that the contention period may start again
• Using PCF automatically sets the NAV, preventing other stations from
sending
• The contention-free period planned initially would have been from t0 to t3
• The point coordinator finished polling earlier, shifting the end of the
contention-free period to t2
• At t4, the cycle starts again with the next super frame
• The transmission properties of the whole wireless network are now
determined by the polling behavior of the access point
• If only PCF is used and polling is distributed evenly, the bandwidth is
also distributed evenly among all polled nodes
• This would resemble a static, centrally controlled time division multiple
access (TDMA) system with time division duplex (TDD) transmission
• This method comes with an overhead if nodes have nothing to send, but
the access point polls them permanently
 802.11 – MAC Frame format
Types
• Control frames, management frames, data frames
Sequence numbers
• important against duplicated frames due to lost ACKs
Addresses
• receiver, transmitter (physical), BSS identifier, sender (logical)
• Miscellaneous
• sending time, checksum, frame control, data
bytes 2 2 6 6 6 2 6 0-2312 4
Frame Duration/ Address Address Address Sequence Address
Data CRC
Control ID 1 2 3 Control 4

bits 2 2 4 1 1 1 1 1 1 1 1
Protocol To From More Power More
Type Subtype Retry WEP Order
version DS DS Frag Mgmt Data
 MAC address format
scenario to DS from address 1 address 2 address 3 address 4
DS
ad-hoc network 0 0 DA SA BSSID -
infrastructure 0 1 DA BSSID SA -
network, from AP
infrastructure 1 0 BSSID SA DA -
network, to AP
infrastructure 1 1 RA TA DA SA
network, within DS

DS: Distribution System

AP: Access Point
DA: Destination Address
SA: Source Address
BSSID: Basic Service Set Identifier
RA: Receiver Address
TA: Transmitter Address
 Special Frames: ACK, RTS, CTS
bytes 2 2 6 4
Frame Receiver
• Acknowledgement Control
Duration
Address
CRC

ACK

bytes 2 2 6 6 4
Frame Receiver Transmitter
RTS Duration CRC
Control Address Address

• Request To Send
bytes 2 2 6 6 4
Frame Receiver Transmitter
CTS Duration CRC
Control Address Address

• Clear To Send
 802.11 - MAC management
• Synchronization
– try to find a LAN, try to stay within a LAN
– timer etc
• Power management
– sleep-mode without missing a message
– periodic sleep, frame buffering, traffic measurements
• Association/Reassociation
– integration into a LAN
– roaming, i.e. change networks by changing access points
– scanning, i.e. active search for a network
• MIB - Management Information Base
– managing, read, write
Synchronization using a Beacon (infrastructure)

beacon interval
(20ms – 1s)

B B B B
access
point
busy busy busy busy
medium
t
value of the timestamp B beacon frame
 Synchronization using a Beacon
(ad-hoc)

beacon interval

B1 B1
station1

B2 B2
station2

busy busy busy busy

medium
t
value of the timestamp B beacon frame random delay
 Power management
• Idea: switch the transceiver off if not needed
• States of a station: sleep and awake
• Timing Synchronization Function (TSF)
– stations wake up at the same time
• Infrastructure
– Traffic Indication Map (TIM)
• list of unicast receivers transmitted by AP
– Delivery Traffic Indication Map (DTIM)
• list of broadcast/multicast receivers transmitted by AP
• Ad-hoc
– Ad-hoc Traffic Indication Map (ATIM)
• announcement of receivers by stations buffering frames
• more complicated - no central AP
• collision of ATIMs possible (scalability?)
• APSD (Automatic Power Save Delivery)
– new method in 802.11e replacing above schemes
 Power management

• Idea: switch the transceiver off if not needed

• For sending device, the device triggers itself. But receiver does not know
in advance. It has to wake up the transceiver periodically
• States of a station: sleep and awake
• If a sender wants to communicate with a power saving station, it has to
buffer data if the station is asleep. The sleeping station has to wake up
periodically and stay awake for certain time.
• During this time the sender announces the destinations of their buffered
data frames. If the station detects that it is a destination of a buffered
packet it has to stay awake until the transmission takes place.
• Timing Synchronization Function (TSF)
– stations wake up at the same time and at the right time
 Power management in infrastructure-based
networks
• Access point buffers all frames destined for stations operating in power
saving mode.
• Along with beacon frame TIM (having a list of stations for which unicast
data frames are buffered in the AP) is sent.
• TSF assures that the sleeping stations will wake up periodically and listen
to the beacon and TIM. If TIM indicates that a unicast frame is buffered for
the station, the station stays awake for the transmission
• To send multicast/broadcast frames, DTIM is used, which is multiple of
TIM intervals.
 Power saving with wake-up patterns
(infrastructure)
TIM interval DTIM interval

D B T T d D B
access
point
busy busy busy busy
medium

p d
station
t
T TIM D DTIM awake

B broadcast/multicast p PS poll d data transmission

to/from the station
Power saving with wake-up patterns (ad-hoc)
ATIM
window beacon interval

B1 A D B1
station1

B2 B2 a d
station2

t
B beacon frame random delay A transmit ATIM D transmit data

awake a acknowledge ATIM d acknowledge data

 802.11 - Roaming
• No or bad connection? Then perform:
• Scanning
-- Active Scanning: Send probes into the medium and wait for an answer
– Passive Scanning: scan the environment, i.e., listen into the medium
for beacon signals
• Re-association Request
– station sends a request to one or several AP(s)
• Re-association Response
– success: AP has answered, station can now participate
– failure: continue scanning
• AP accepts Reassociation Request
– signal the new station to the distribution system
– the distribution system updates its data base (i.e., location information)
– typically, the distribution system now informs the old AP so it can
release resources
• Fast roaming – 802.11r
– e.g. for vehicle-to-roadside networks