Medium Access Control (MAC) Techniques
5.2.1 Introduction
A network of computers based on multi-access medium requires a protocol for effective
sharing of the media. As only one node can send or transmit signal at a time using the
broadcast mode, the main problem here is how different nodes get control of the medium
to send data, that is “who goes next?”. The protocols used for this purpose are known as
Medium Access Control (MAC) techniques. The key issues involved here are - Where and
How the control is exercised.
‘Where’ refers to whether the control is exercised in a centralised or distributed
manner. In a centralised system a master node grants access of the medium to other
nodes. A centralized scheme has a number of advantages as mentioned below:
o Greater control to provide features like priority, overrides, and guaranteed
bandwidth.
o Simpler logic at each node.
o Easy coordination.
Although this approach is easier to implement, it is vulnerable to the failure of the
master node and reduces efficiency. On the other hand, in a distributed approach all the
nodes collectively perform a medium access control function and dynamically decide
which node to be granted access. This approach is more reliable than the former one.
‘How’ refers to in what manner the control is exercised. It is constrained by the
topology and trade off between cost-performance and complexity. Various approaches
for medium access control are shown in Fig. 5.2.1. The MAC techniques can be broadly
divided into four categories; Contention-based, Round-Robin, Reservation-based and.
Channelization-based. Under these four broad categories there are specific techniques, as
shown in Fig. 5.2.1. In this lesson we shall concentrate of the MACs of the first two
categories, which have been used in the legacy LANs of the IEEE standard. The
CSMA/CA, a collision-free protocol used in wireless LAN, will be presented in Lesson
5.5. Channalization-based MACs, which are used in cellular telephone networks, will be
covered in Lesson 5.6. And the reservation-based MACs, which are used in satellite
networks, will be discussed in Lesson 5.7.
Figure 5.2.1 Possible MAC techniques
5.2.2 Goals of MACs
Medium Access Control techniques are designed with the following goals in mind.
x Initialisation: The technique enables network stations, upon power-up, to enter
x
the state required for operation.
Fairness: The technique should treat each station fairly in terms of the time it is
made to wait until it gains entry to the network, access time and the time it is
x
allowed to spend for transmission.
Priority: In managing access and communications time, the technique should be
able to give priority to some stations over other stations to facilitate different type
x
of services needed.
Limitations to one station: The techniques should allow transmission by one
x
station at a time.
Receipt: The technique should ensure that message packets are actually received
(no lost packets) and delivered only once (no duplicate packets), and are received
x
in the proper order.
Error Limitation: The method should be capable of encompassing an
x
appropriate error detection scheme.
Recovery: If two packets collide (are present on the network at the same time), or
if notice of a collision appears, the method should be able to recover, i.e. be able
to halt all the transmissions and select one station to retransmit.
x Reconfigurability: The technique should enable a network to accommodate the
addition or deletion of a station with no more than a noise transient from which
x
the network station can recover.
Compatibility: The technique should accommodate equipment from all vendors
x
who build to its specification.
Reliability: The technique should enable a network to confine operating inspite of
a failure of one or several stations.
5.2.3 Round Robin Techniques
In Round Robin techniques, each and every node is given the chance to send or transmit
by rotation. When a node gets its turn to send, it may either decline to send, if it has no
data or may send if it has got data to send. After getting the opportunity to send, it must
relinquish its turn after some maximum period of time. The right to send then passes to
the next node based on a predetermined logical sequence. The right to send may be
controlled in a centralised or distributed manner. Polling is an example of centralised
control and token passing is an example of distributed control as discussed below.
5.2.3.1 Polling
The mechanism of polling is similar to the roll-call performed in a classroom. Just like
the teacher, a controller sends a message to each node in turn. The message contains the
address of the node being selected for granting access. Although all nodes receive the
message, only the addressed node responds and then it sends data, if any. If there is no
data, usually a “poll reject” message is sent back. In this way, each node is interrogated
in a round-robin fashion, one after the other, for granting access to the medium. The first
node is again polled when the controller finishes with the remaining codes.
The polling scheme has the flexibility of either giving equal access to all the
nodes, or some nodes may be given higher priority than others. In other words, priority of
access can be easily implemented.
Figure 5.2.2 Polling using a central controller
Polling can be done using a central controller, which may use a frequency band to
send outbound messages as shown in Fig. 5.2.2. Other stations share a different
frequency to send inbound messages. The technique is called frequency-division duplex
approach (FDD). Main drawbacks of the polling scheme are high overhead of the polling
messages and high dependence on the reliability of the controller.
Polling can also be accomplished without a central controller. Here, all stations
receive signals from other stations as shown in Fig. 5.2.3. Stations develop a polling
order list, using some protocol.
Figure 5.2.2 Polling in a distributed manner
5.2.3.2 Token Passing
In token passing scheme, all stations are logically connected in the form of a ring and
control of the access to the medium is performed using a token. A token is a special bit
pattern or a small packet, usually several bits in length, which circulate from node to
node. Token passing can be used with both broadcast (token bus) and sequentially
connected (token ring) type of networks with some variation in the details as considered
in the next lesson.
In case of token ring, token is passed from a node to the physically adjacent node.
On the other hand, in the token bus, token is passed with the help of the address of the
nodes, which form a logical ring. In either case a node currently holding the token has the
‘right to transmit’. When it has got data to send, it removes the token and transmits the
data and then forwards the token to the next logical or physical node in the ring. If a node
currently holding the token has no data to send, it simply forwards the token to the next
node. The token passing scheme is efficient compared to the polling technique, but it
relies on the correct and reliable operation of all the nodes. There exists a number of
potential problems, such as lost token, duplicate token, and insertion of a node, removal
of a node, which must be tackled for correct and reliable operation of this scheme.
Figure 5.2.3 A token ring network
Figure 5.2.4 Token passing mechanism
Performance: Performance of a token ring network can be represented by two
parameters; throughput, which is a measure of the successful traffic, and delay, which is
a measure of time between when a packet is ready and when it is delivered. A station
starts sending a packet at t = t0, completes transmission at t = t0 + a, receives the tail at t0
+ 1 + a. So, the average time (delay) required to send a token to the next station = a/N.
and throughput, S = 1/(1 + a/N) for a<1 and S = 1/a(1 + 1/N) for a>1.
5.2.4 Contention-based Approaches
Round-Robin techniques work efficiently when majority of the stations have data to send
most of the time. But, in situations where only a few nodes have data to send for brief
periods of time, Round-Robin techniques are unsuitable. Contention techniques are
suitable for bursty nature of traffic. In contention techniques, there is no centralised
control and when a node has data to send, it contends for gaining control of the medium.
The principle advantage of contention techniques is their simplicity. They can be easily
implemented in each node. The techniques work efficiently under light to moderate load,
but performance rapidly falls under heavy load.
5.2.4.1 ALOHA
The ALOHA scheme was invented by Abramson in 1970 for a packet radio network
connecting remote stations to a central computer and various data terminals at the campus
of the university of Hawaii. A simplified situation is shown in Fig. 5.2.5. Users are
allowed random access of the central computer through a common radio frequency band
f1 and the computer centre broadcasts all received signals on a different frequency band
f2. This enables the users to monitor packet collisions, if any. The protocol followed by
the users is simplest; whenever a node has a packet to sent, it simply does so. The
scheme, known as Pure ALOHA, is truly a free-for-all scheme. Of course, frames will
suffer collision and colliding frames will be destroyed. By monitoring the signal sent by
the central computer, after the maximum round-trip propagation time, an user comes to
know whether the packet sent by him has suffered a collision or not.
Central
f1 --random access
Computer f2 --broadcast
f2 f2
f2 f2
f1
D
f1
f1
f1
A
B C
Figure 5.2.5 Simplified ALOHA scheme for a packet radio system
Figure 5.2.6 Collision in Pure ALOHA
It may be noted that if all packets have a fixed duration of W (shown as F in Figure
time from W before to until W after the start of the packet A as shown in Fig. 5.2.6. This
5.2.7), then a given packet A will suffer collision if another user starts to transmit at any
gives a vulnerable period of 2W. Based on this assumption, the channel utilization can be
computed. The channel utilization, expressed as throughput S, in terms of the offered
load G is given by S=Ge-2G.
Figure 5.2.7 Vulnerable period in Pure ALOHA
Based on this, the best channel utilisation of 18% can be obtained at 50 percent of
the offered load as shown in Fig. 5.2.8. At smaller offered load, channel capacity is
underused and at higher offered load too many collisions occur reducing the throughput.
The result is not encouraging, but for such a simple scheme high throughput was also not
expected.
Figure 5.2.8 Throughput versus offered load for ALOHA protocol
Figure 5.2.9 Slotted ALOHA: Single active node can continuously transmit at full rate of
channel
Subsequently, in a new scheme, known as Slotted ALOHA, was suggested to
slots equal to W and packet transmission can start only at the beginning of a slot as shown
improve upon the efficiency of pure ALOHA. In this scheme, the channel is divided into
in Fig. 5.2.9. This reduces the vulnerable period from 2W to W and improves efficiency by
reducing the probability of collision as shown in Fig. 5.2.10. This gives a maximum
throughput of 37% at 100 percent of offered load, as shown in Figure 5.2.8.
Figure 5.2.10 Collision in Slotted ALOHA
5.2.5 CSMA
The poor efficiency of the ALOHA scheme can be attributed to the fact that a node start
transmission without paying any attention to what others are doing. In situations where
propagation delay of the signal between two nodes is small compared to the transmission
time of a packet, all other nodes will know very quickly when a node starts transmission.
This observation is the basis of the carrier-sense multiple-access (CSMA) protocol. In
this scheme, a node having data to transmit first listens to the medium to check whether
another transmission is in progress or not. The node starts sending only when the channel
is free, that is there is no carrier. That is why the scheme is also known as listen-before-
talk. There are three variations of this basic scheme as outlined below.
(i) 1-persistent CSMA: In this case, a node having data to send, start sending, if the
channel is sensed free. If the medium is busy, the node continues to monitor until the
channel is idle. Then it starts sending data.
(ii) Non-persistent CSMA: If the channel is sensed free, the node starts sending the
packet. Otherwise, the node waits for a random amount of time and then monitors the
channel.
(iii) p-persistent CSMA: If the channel is free, a node starts sending the packet. Otherwise
the node continues to monitor until the channel is free and then it sends with probability
p.
The efficiency of CSMA scheme depends on the propagation delay, which is represented
by a parameter a, as defined below:
Propagation delay
a = Packet transmission time.
The throughput of 1-persistent CSMA scheme is shown in Fig. 5.2.11 for different values
of a. It may be noted that smaller the value of propagation delay, lower is the vulnerable
period and higher is the efficiency.
5.2.6 CSMA/CD
CSMA/CD protocol can be considered as a refinement over the CSMA scheme. It has
evolved to overcome one glaring inefficiency of CSMA. In CSMA scheme, when two
packets collide the channel remains unutilized for the entire duration of transmission time
of both the packets. If the propagation time is small (which is usually the case) compared
to the packet transmission time, wasted channel capacity can be considerable. This
wastage of channel capacity can be reduced if the nodes continue to monitor the channel
while transmitting a packet and immediately cease transmission when collision is
detected. This refined scheme is known as Carrier Sensed Multiple Access with Collision
Detection (CSMA/CD) or Listen-While-Talk.
On top of the CSMA, the following rules are added to convert it into CSMA/CD:
(i) If a collision is detected during transmission of a packet, the node immediately ceases
transmission and it transmits jamming signal for a brief duration to ensure that all stations
know that collision has occurred.
(ii) After transmitting the jamming signal, the node waits for a random amount of time
and then transmission is resumed.
The random delay ensures that the nodes, which were involved in the collision are
not likely to have a collision at the time of retransmissions. To achieve stability in the
back off scheme, a technique known as binary exponential back off is used. A node will
attempt to transmit repeatedly in the face of repeated collisions, but after each collision,
the mean value of the random delay is doubled. After 15 retries (excluding the original
try), the unlucky packet is discarded and the node reports an error. A flowchart
representing the binary exponential back off algorithm is given in Fig. 5.2.11.
Performance Comparisons: The throughput of the three contention based schemes with
respect to the offered load is given in Fig 5.2.12. The figure shows that pure ALHOA
gives a maximum throughput of only 18 percent and is suitable only for very low offered
load. The slotted ALHOA gives a modest improvement over pure ALHOA with a
maximum throughput of 36 percent. Non persistent CSMA gives a better throughput than
1-persistent CSMA because of smaller probability of collision for the retransmitted
packets. The non-persistent CSMA/CD provides a high throughput and can tolerate a
very heavy offered load. Figure 5.2.13 provides a plot of the offered load versus
throughput for the value of a = 0.01.
Figure 5.2.11 Binary exponential back off algorithm used in CSMA/CD
Figure 5.2.12 Comparison of the throughputs for the contention-based MACs
Figure 5.2.13 A plot of the offered load versus throughput for the value of a = 0.01
Performance Comparison between CSMA/CD and Token ring: It has been observed
that smaller the mean packet length, the higher the maximum mean throughput rate for
token passing compared to that of CSMA/CD. The token ring is also least sensitive to
workload and propagation effects compared to CSMS/CD protocol. The CSMA/CD has
the shortest delay under light load conditions, but is most sensitive to variations to load,
particularly when the load is heavy. In CSMA/CD, the delay is not deterministic and a
packet may be dropped after fifteen collisions based on binary exponential back off
algorithm. As a consequence, CSMA/CD is not suitable for real-time traffic.
Fill In The Blanks:
1. The basic question which has to be answered by the medium-access control
techniques is “How Goes ________”?
2. In _______________ technique, each node gets a chance to access the medium by
rotation.
3. The key issues involved in MAC protocol are - Where and ______ the control is
exercised.
4. ‘Where’ refers to whether the control is exercised in a _____________ or
______________ manner.
5. The _______________ techniques can be broadly categorized into three types;
Round-Robin, Reservation and___________.
6. ______________ is an example of centralized control and ______________ is an
example of distributed control
7. In Polling technique, if there is no data, usually a ______________ message is
sent back.
8. In pure ALOHA, channel utilization, expressed as throughput S, in terms of the
offered load G is given by _________
9. In slotted ALOHA, a maximum throughput of ________ percent at 100 percent of
offered load can be achieved, while it is ______ percentage for pure ALOHA.
10. _________________ is abbreviated as CSMA/CD and is also known as
__________.
11. To achieve stability in CSMA/CD back off scheme, a technique known as
______________ is used
Solutions:
1. Next
2. token passing
3. How
4. centralized, distributed
5. asynchronous, Contention
6. Polling, token passing
7. poll reject
8. S=Ge-2G.
9. 37, 18
10. Carrier Sensed Multiple Access with Collision Detection, Listen-While-Talk .
11. binary exponential back off
Short Answer Questions:
Q-1. In what situations contention based MAC protocols are suitable?
Ans: Contention based MAC protocols are suitable for bursty nature of traffic under
light to moderate load. These techniques are always decentralized, simple and easy to
implement.
Q-2. What is vulnerable period? How it affects the performance in MAC protocols?
period. Let, all packets have a fixed duration O. Then vulnerable period is 2O in pure
Ans: The total period of time when collision may occur for a packet is called vulnerable
ALOHA scheme and O in slotted ALOHA scheme. If vulnerable period is long,
probability of the occurrence collision increases leading to reduction in throughput.
Q-3. How throughput is improved in slotted ALOHA over pure ALOHA?
Ans: In pure ALOHA vulnerable period is 2O.
So, S/G = e-2G or throughput S = G e-2G , where G is the total number of packets.
Maximum value of G = 0.5 or maximum throughput Smax = 1/2e.
In slotted ALOHA, vulnerable period is O and S/G = e-G or throughput S = G e-G . Here,
maximum value of G is 1 and maximum throughput Smax = 1/e.
Q-4. What is the parameter ‘a’? How does it affect the performance of the CSMA
protocol?
Ans: The efficiency of CSMA scheme depends on propagation delay, which is
represented by a parameter ‘a’ as defined below.
propagation delay
a = --------------------------------
packet transmission time
Smaller the value of propagation delay, lower is the vulnerable period and higher
is the efficiency. If propagation delay is zero, collision cannot occur in CSMA scheme.
But in practice, there is some delay and depending on the value of ‘a’ collision occurs.
Q-5. How performance is improved in CSMA/CD protocol compared to CSMA
protocol?
Ans: In CSMA scheme, a station monitors the channel before sending a packet.
Whenever a collision is detected, it does not stop transmission leading to some wastage
of time. On the other hand, in CSMA/CD scheme, whenever a station detects a collision,
it sends a jamming signal by which other station comes to know that a collision occurs.
As a result, wastage of time is reduced leading to improvement in performance.