shanmu-conference paper

12

Click here to load reader

Upload: shanmugasundaram-rk

Post on 18-Aug-2015

23 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Shanmu-CONFERENCE PAPER

AN EFFECTIVE BACKOFF ALGORITHM FOR

COLLISION REDUCTION IN MANET

R.K. Shanmugasundaram, S. Vaidheeswaran

BTECH, B.S Abdur Rahman University, Vandalur, Tamilnadu, (India)

Professor, Dr.N. Sabiyath Fatima, Dept of CSE, B.S Abdur Rahman University,Vandalur, Tamilnadu, (India)

ABSTRACT

In Mobile Ad-hoc Networks (MANETS), every node performs both as transmitter and receiver. The existing

backoff models do not accurately predict the performance of the wireless network. Also, the existing models

suffer with high packet collisions. Whenever a collision occurs, the contention window (CW) of the station is

doubled until it reaches the maximum value. The main objective of this project is to reduce collision using

contention window Multiplicative Increase Decrease Backoff (CWMIDB) scheme. The purpose of increasing

CW is to reduce the collision probability by distributing the traffic into a larger time space. In wireless Ad hoc

networks, the CWMIDB algorithm dynamically controls the contention window of the nodes experiencing

collisions. During packet transmission, the backoff counter is uniformly selected from the given range of [0,

CW-1]. Here, CW is known as contention window and its value depends on the number of failed transmissions

for the packet. At the first transmission attempt, CW is set to minimum value (Cmin), if transmission attempt fails

then its value gets doubled, and again set to minimum value on successful transmission. CWMIDB is simulated

in NS2 environment and its performance is compared with Binary Exponential Backoff Algorithm. The

simulation results shows improvement in transmission probability compared to that of the existing backoff

algorithm.

Keywords— backoff, MANET, contention window,CWMIDB.

I. INTRODUCTİON

Mobile Ad Network (MANET) is an infrastructure-less dynamic network without fixed routers. All nodes are

capable of moving and can be connected dynamically. The responsibilities for organizing and controlling the

network are distributed among the terminals themselves. The entire network is mobile, in this type of network,

some pairs of terminals may not be able to communicate directly with each other and relaying of some messages

is required so that they are delivered to their destinations. The main drawback of MANET is energy

conservation. Energy consumption increases due to packet loss. This packet loss occurs due to collision between

the nodes during transmission of data. For wireless networks, the devices operating on battery try to pursue the

energy efficiently by reducing the energy they consume, while data transmission. The features of MANET

introduce several challenges. These include, collision due to transmission errors. This system breaks down when

two computers attempt to transmit at the same time. This is a case of collision. To avoid collision, carrier

sensing mechanism is used. Here each computer listens to the network before attempting to transmit. If the

network is busy, it waits until network quiets down. In carrier detection, computers continue to listen to the

Page 2: Shanmu-CONFERENCE PAPER

network as they transmit. If computer detects another signal that interferes with the signal it is sending, it stops

transmitting. Both computers then wait for random amount of time and attempt to transmit. A collision can be

detected by listening to the shared medium immediately after transmitting and identifying collision

characteristics; or by capturing data from the medium and performing error detection.An alternative method to

handle collisions in a contention based system is to attempt to avoid them. Some systems may utilize a strict

scheduling guideline to identify who may use which resources another system may have the senders listen to the

channel immediately prior to transmitting and determine suitable times to transmit. The purpose of using

CWMIDB algorithm in IEEE 802.11 is to minimize collisions during contention between multiple nodes and

also in the presence of hidden nodes. This is a very simple backoff algorithm in which the size of the contention

window increases and decreases rather than increasing exponentially. The basic operation of the MAC protocol

is by successive Frame Exchange Sequences (FES) RTS-CTS-DATA-ACK as in 802.11 DCF. Each transmitter

before attempting to transmit must find the carrier to be idle for a time period of Distributed Inter-Frame

Spacing (DIFS) seconds. After deferring for the DIFS period the station selects a back off value for an

additional deferral time before transmitting. This back off period corresponds to an integer number of time slots

of the protocol and is selected based on Contention Window Multiplicative Increase Decrease Backoff

Algorithm. When some other transmission is detected the back off procedure pauses and starts over when the

medium is found to be idle for one DIFS period. If collision occurs, then the stations that were involved have to

restart the access procedure with DIFS period and new back off value.

II. AN EFFECTIVE BACKOFF ALGORITHM FOR COLLISION REDUCTION IN

MANET

The major problem in MANET is degradation in throughput because of node mobility, unreliable medium,

interference and route failure. The degradation in throughput is because MANET is unable to distinguish packet

loss. The packet loss occurs due to hidden node and congestion. This problem can be solved by using certain

backoff scheme. The backoff algorithm is a part of Media Access Control (MAC) protocol which is used to

avoid collision in Mobile Ad hoc Network (MANET). When the nodes in the network try to access the channel,

one of these nodes gains access the channel while the other nodes still contend for a time period. Collision

occurs in mobile ad-hoc networks when the node chooses the same value for data transmission. In Figure 3.1, A

and B are two nodes. During the transmission of data between these two nodes, collision occurs. After receiving

the collision signal, transmission is stopped from A and B. They will try to retransmit, since there is no

restriction.

Figure 2.1 Collision Problem

Page 3: Shanmu-CONFERENCE PAPER

To overcome from the problem of collision, a new Contention Window Multiplicative Increase Decrease

Backoff scheme is proposed. CWMIDB provides minimum transmission delay and maximum throughput by

improving collision. This collision can be improved by certain mechanism. In the existing Binary Exponential

Backoff algorithm, for each packet transmission, the backoff time is uniformly chosen in the range (0, W-1)

where W is the initial contention window size. When the packet transmission is not successful the node

increases the contention window size, so the node has to sense the channel for long period of time. Due to this

throughput of the wireless network decreases automatically. The average delay required for the packet delivery

increases.

Figure 2.2 Backoff Concept

The Contention Window Multiplicative Increase Decrease Backoff Algorithm (CWMIDB) consist of three

phases, they are Identification of channel idleness, Setting up Backoff Counter, Performance comparison of

CWMIDB & Binary Exponential Backoff Algorithm. The steps for Contention Window Multiplicative Increase

Decrease Backoff Algorithm is implemented as follows,

CW = min [4*CW, CWmax] --------- (2.1)

The equation 2.2 describes that size of the contention window increase by four times during collision

CW = CWmin ------------------ (2.2)

The equation 2.2 describes that size of the contention window decrease during successful transmission

The CWMIDB algorithm steps are discussed

Step 1: Initially source node sends data when channel is in idle state until the destination is reached. The data

transmission process consists of two states namely idle and non-idle state.

Step 2: In Figure 2.3 S,A,B,D are nodes where S is source node and D is destination node, This node check for

idleness, if channel is in idle state, then it transfers the data to destination. If channel is not idle then it waits for

random backoff time.

Figure 2.3 A Scenario of Collision Detection

Page 4: Shanmu-CONFERENCE PAPER

Step 3: When random backoff time value reaches zero it will transmit packet again to the destination.

Step 4: If the frame is correctly received, the receiving node sends an acknowledgment (ACK) frame to sender.

If ACK is not received, the size of the contention window size increases and retransmit packet again to the

destination.

Step 4.1: During successful transmission, the size of the contention window decreases to minimum.

Step 4.2: During the second transmission attempt, the size of contention window increases exponentially by CW

= CW x 4.

Step 5: During third transmission attempt, the size of the contention window decreases by CW = current value

of CW / 2

Step 6: If acknowledgement is received, go to step1.

Step 7: Stop

III Pseudocode for CWMIDB algorithm during successful transmission

Step 1: Set BackOff Timer to initial value

Step 2: While BackOff Timer ≠ 0 do

For each Time Slot

If channel is idle then BackOff Timer = BackOff Timer – 1

If channel is idle for more than DIFS then

Send

If (Send Failure) then

If (NumberOfBackoffs<=W) then

CW= CW * K

Else if (NumberOfBackoffs>=W)

CW= CW / K

BackOffTimer = Random x; 1 ≤ x ≤ CW -1

Else

CW = CW - 1

BackOffTimer = 0

Go to Step 1

Step 3: Stop

[Where

DIFS – Distributed Inter-Frame Spacing

CW-Contention Window]

/* Channel idleness process*/

When idle

MinCW=32;

MaxCW=256;

Backoff range = [0,CW];

/*During Successful Transmission*/

MinCW=32;

MaxCW=256, 1024;

Backoff range = [0,CW];

Page 5: Shanmu-CONFERENCE PAPER

if (success) node's CW = MinCW;

node's BC = randomly select (0, MinCW);

end if

/*During Packet Collision*/

MinCW=32;

MaxCW=256, 1024;

Backoff range = [0,CW];

if (collision),

node's CW =

min(4*currentCW, MaxCW);

node's BC = randomly

select (0, newCW);

end

3.1 Channel Idleness phase

In channel idleness phase every node must sense the medium, in order to check the state of the channel (idle or

busy). If a node has data to send, but it sees that the channel is busy, then it waits for the end of transmission. At

the end of transmission, it must wait for a DIFS. The back off time counter continuously gets decremented until

the channel is sensed in an “Idle” state. It goes in “Frozen” State when a transmission is detected on the channel

and it goes to the reactivated state when the channel is sensed idle again for more than a DFIS. As soon as the

back-off time reaches zero the station starts the transmission. If two or more wireless nodes finish their

countdowns at the same time-slot, there occurs a collision between RTS (ready to send) packets if the

CSMA/CA (carrier sense multiple access with collision avoidance) is implemented, otherwise two data packets

collide with each other. If there is a collision, every node which participated in the collision multiplies its

contention window by the multiplicative factor m.

Figure 3.1 IDLE Procedure Finite State Representation

In Figure 3.1, the contention period can be determined by the CWMIDB algorithm. It increments the appropriate

retry counter associated with the frame. The CWMID Backoff mechanism chooses a random number which

represents the amount of time must elapse while there are not any transmissions, i.e., the medium is idle before

the listening station may attempt to begin its transmission again. The random number resulting from this

algorithm is uniformly distributed in a range, called the contention window, the size of which increases with

every attempt to transmit that is deferred, until a maximum size is reached for the range. Once a transmission is

successfully completed, the range is reduced to its minimum value for the Next transmission.

Page 6: Shanmu-CONFERENCE PAPER

3.2 Backoff Counter Phase

In backoff counter phase, it attempts to double the contention window size during collision .Set the backoff counter

(BOC) between 0 and CWlow. Decrement the BOC by 1 if the channel is idle. If BOC reaches zero, transmit the data

frame. If acknowledgment is received then, transmit the next data frame. If the acknowledgement is not received

select the BOC between 0 and CW in the next backoff stage as

For second transmission attempt, CW = CW x 4. For third transmission attempt, CW = current value of CW / 2. For

fourth transmission attempt, CW = current value of CW X 4. For fifth transmission attempt, CW = current value of

CW / 2 and so on. If contention window is high, it will select the BOC between 0 and CW in the current backoff

stage.

Figure 3.2 BACKOFF Procedure Finite State Representation.

The Figure 3.2 depicts the backoff procedure for finite state representation.To prevent all nodes sensing the

channel from beginning of the transmission at the same time, each node chooses a random waiting time called

Backoff before starting transmission. The range from which the Backoff value is chosen is called the Contention

Window (CW). If the channel becomes busy during Backoff, Backoff counter decrementing is stopped until the

channel becomes idle again. For example, let us assume that the current backoff stage is ‘i’ with contention

windowCW( i ) = 4i * CWmin , and there is a successful transmission, the next backoff stage will be stage 0

with contention window CW( 0 ) = 31 according to the specification. But if the number of competing nodes is

large enough (>>31), the new collision will likely occur at the backoff stage 0. The main argument is that since

the current backoff stage is ‘i’ some collision must have occurred recently at the previous stage. Now if the

number of current competing nodes is larger than or close to CW( i ), and if the backoff stage is set to 0, there is

a high probability that new Collision will happen. So resetting the contention window after every successful

transmission is an inefficient approach if the number of nodes is large.

3.3 PERFORMANCE COMPARISON OF CWMIDB AND BEB

Here in this project performance comparison is done with CWMIDB and Binary Exponential Backoff (BEB) .

This performance comparison is done by taking into the account the performance metrics like packet delivery

ratio (PDR), Energy, Throughput. While making comparison CWMIDB is proven to be highly energy

conservation and efficient one than other.

Page 7: Shanmu-CONFERENCE PAPER

The characteristics of NS2 parameter like throughput, packet information, etc can be plotted using trace graph.

The following Table 7.2 illustrates the comparison of various performance metrices like Packet Delivery Ratio

(PDR), Energy and throughput. These metrices are calculated by taking in to the account the number of sending

packets, number of receiving packets at destination and considerably the number of routing packets which is

generated in the trace file. The packet delivery ratio (PDR) of CWMIDB gradually increases from 10.28 percent

to 91.92 percent. The PDR ratio is high because CWMIDB send the packets by increasing and decreasing the

contention window size. CWMIDB initially increases and then decreases. The simulation analysis of CWMIDB

algorithm is obtained by considering the following performance metrics like, Packet delivery ratio and

Throughput

IV Packet Delivery Ratio (PDR)

Packet Delivery Ratio (PDR) is the ratio between the number of packets transmitted by a traffic source and the

number of packets received by a traffic sink. It measures the loss rate as seen by transport protocols and as such,

it characterizes both the correctness and efficiency of ad hoc routing protocols. In CWMIDB the packet

Delivery Ratio is high as 97.94%.

PDR is defined as the total amount of packets received by the receiver and amount of data packet sent by source.

The formula to calculate the packet delivery ratio is defined as follows,

PDR=[ ]* 100

Then by taking into the account number of sending packets, receiving packets and routing packets the graph is

plotted. Here, the packet delivery ratio gradually increases with time in CWMIDB. Here, the PDR of CWMIDB

is 97.94%.The Figure 4.1 shows the packet delivery ratio between CWMIDB and BEB.

Figure 4.1 Packet delivery ratio between CWMIDB and BEB

V Throughput analysis

It is the ratio of the total amount of data that reaches a receiver from a sender to the time it takes for the

receiver to get the last packet. When comparing the routing throughput by each of the node, AODV has the high

throughput. It measures of effectiveness of a routing protocol. The throughput value of CWMIDB algorithm is

high when compared to BEB. Here, the throughput ratio increases with increase in number of packets received.

Page 8: Shanmu-CONFERENCE PAPER

Figure 4.2 Throughput comparison between CWMIDB and BEB

The Figure 7.2 shows the throughput comparison between CWMIDB and BEB. It shows that

CWMIDB has high throughput than BEB. The throughput percentage of CWMIDB is 97.14%

VI CONCLUSION AND FUTURE WORK

In this project, Contention Window based Multiplicative Increase Decrease backoff (CWMIDB) algorithm for

IEEE 802.11 based wireless networks is proposed. The main feature of this algorithm is to avoid channel

capture and to decrease the number of collisions. In this algorithm, the size of the contention window will not be

doubled immediately following a collision. Rather it increases by four times for first collision and decreases by

half for second collision and so on. The performance of the CWMIDB algorithm is very efficient especially

when basic access mechanism is employed under high congested environments as well as in Ad hoc networks.

The simulation result and analysis shows that Contention Window Multiplicative Increase Decrease Backoff

algorithm has higher efficiency with high packet delivery ratio and throughput than the Binary Exponential

Backoff algorithm.

CWMIDB can be extended further by taking into account the data transmission among number of nodes. If

number of node increases there is a slight deflection in the performance. The future work concentrates on the

reduction of collision by increasing the number of nodes. Integration of modified BEB and CWMIDB may

produce high throughput which may help in increasing the efficiency of CWMIDB.

REFERENCES

[1] Tatineni Madhavi, Gottapu SasiBhushanaRao, Development of Collision Alleviating DCF Protocol with

Efficient Backoff Algorithm for Wireless Ad hoc Networks. November 2014, 80:1791–1814.1.

[2] Sasi BhushanaRao, G., &Madhavi, T, Performance analysis of collision alleviating distributed coordination

function protocol in congested wireless networks - a Markov chain analysis, IET Networks, 2013, 2(4), 204–

213.

[3] Ibrahim Sayed Ahmad., Modified Binary Exponential Backoff Algorithm to Minimize Mobiles

Communication Time”, I.J. Information Technology and Computer Science, 2014, 20-29,.

[4] Xiangyu, P., Letian, J.,&Guozhi,X. Performance analysis of hybrid backoff algorithm of wireless

LAN.InProceedings of 5th international conference on WiCom, 2007, pp. 1853–1856.

[5] Sartthong, J., &Sittichivapak, S. Backoff algorithm optimization for IEEE802.11 wireless local area

networks. In Proceedings of ECTI-CON, 2012, pp. 1–4.

[6] Fan, J., Gao, F.,Wang,W. S., & Dong, G. F. Performance analysis of an adaptive backoff scheme for ad hoc

networks. In Proceedings of IEEE international conference on CIT, 2008, pp. 624–629.

[7] Albalt, M.,&Nasir, Q. Adaptive backoff algorithm for IEEE 802.11MAC protocol.International Journal of

Communications, Network and System Sciences, 2009, 2(4), 300–317.

Page 9: Shanmu-CONFERENCE PAPER

[8] Babich, F.,&Comisso, M. Throughput and delay analysis of 802.11-based wireless networks using smart and

directional antennas. IEEE Transactions on Communications, 57(5), 2009, 1413–1423.

[9] Zhalehpoor, S., &ShahriarShahhoseini, H. SBA backoff algorithm to enhance the quality of service in

MANETs. In Proceedings of international conference on signal acquisition and processing, 2009, pp. 43–47.

[10] Chatzimisios, P., Boucouvalas, A., & Vitsas, V. Performance analysis of IEEE 802.11 DCF in presence of

transmission errors. In Proceedings of IEEE international conference on communications (7), 2004, pp. 3854–

3858.

[11] Fan, J., Gao, F.,Wang,W. S., & Dong, G. F. Performance analysis of an adaptive backoff scheme for ad hoc

networks. In Proceedings of IEEE international conference on CIT, 2008, pp. 624–629.

[12] Peng Xiangyu, Hybrid Backoff Algorithm based on DCF protocol. IEEE international conference on 2007,

1-4244-1312-5.