Record Adaptive Contention. Window for IEEE 802.11 wireless networks. Sakshi Bhandari,Anuradha deptt of CS Engg and IT. ITM University. Gurgaon,India.
2013 International Conference on Machine Intelligence and Research Research and Advancement
Record Adaptive Contention Window for IEEE 802.11 wireless networks
Sakshi Bhandari,Anuradha
Prabhjot Kaur
deptt of CS Engg and IT ITM University Gurgaon,India
deptt of EEC Engg ITM University Gurgaon,India
incremented by 1 to minimize the probabilitiy of collision.Back off value has to be from 2^CW. After every data transmission,sender waits for fixed amount of time to receive the acknowledgement packet ,ACK.Otherwise,whole of the above mentioned procedure for setting wait timer is to be followed.At the receiver side,station senses the channel to be idle for SIFS amount of time interval.Since SIFS is lesser amount of wait interval than DIFS,Therefore,no other station is able to sense channel free for transmission until the completion of sending of ACK.In the above mechanism ,the collision avoidance issue largely depends on back off counter value.QoS of MAC layer,Hence is dependent on the efficiency of backoff algorithm.In binary exponential algorithm(BEB),there is keeping the track of previous transmissions of the station.This aspect of information is used in our proposal.
Abstract— IEEE 802.11 specification is used in deployment of WLANs.Medium Access Control of Data link layer uses mechanism called DCF to access the wireless channel.Among the major QoS issues in DCF is the lack of analyzing the current network situation for adjusting the contention window for back off procedure.In this paper,we present a novel algorithm for computing the value of contention window size.The algorithm proposed gives a increased throughput and PDR over the traditional Binary Exponential Back off (BEB) of DCF function. Keywords- IEEE 802.11,BEB,Record Adaptive Contention Window.
I.
INTRODUCTION
Recent scenario of extensive use of wireless communication has led to the increased interest in improving the efficiency of WLAN techniques.IEEE 802.11 is one of the extensively used specification for PHY and MAC layers of wireless lan.This standard provides the function DCF for MAC .DCF has been analyzed in [4],[5] by various researchers to overcome its shortcomings and increase its throughput .Many have come up various with algorithms and approaches to replace or modify the binary exponential algorithm(BEB) of DCF. In this paper,we have presented a mechanism called record adaptive contention window(RACW) attempting to use the record of earlier transmissions in determining the value for contention window in back off procedure.The paper is organized as follows.Section i describes the DCF scheme briefly.Section ii gives introduction about the literature of the work done previously.section iii explains our proposed mechanism. Section iv deals with simulation and analysis of the record adaptive contention window algorithm.section v concludes our paper. II.
III.
Several researchers have contributed to the improvement of throughput performance of IEEE 802.11 MAC layer.Various designs are proposed for algorithm for setting of contention window for back off procedure.In the literature work done on analysis of MAC layer,good number of proposals could be defined as the ones using a predefined scale to set the contention window value after success or failure of transmitted data packet.Here, Author in [1] has described contention window mechanism which uses a fixed scale for defining the value for contention window.Advantage of the proposal lie in decrese in computations to be performed. In [3] author has tried to improve the MAC layer efficiency by differentiating between the unsuccessful transmissions due to frame collisions and due to error in wireless medium.The role of contention window is to avoid collision between various data frames .but 802.11 doesn’t differentiate between frame loss reasons,thus,changes the contention window size every time frame loss occurs work presented by bianchi in [4],[5] on the performance analysis of DCF using markov model has led to the indept mathematical analyses of DCF which has been extensively used in various proposals. Separate category of work presented by the author in [6] increases the efficiency of traditional DCF by manipulating the contention window
DISTRIBUTED COORDINATED FUNCTION
DCF is fundamentally based on CSMA/CA protocol.In this approach,whenever a node has a data packet to send it senses the channel to be idle for DIFS period.In effect it generates a back off timer choosen from the interval [0,CW1],where CW signifies contention window. Initial value of cw is cwmin and goes upto cwmax .both these are predefined. After every unsuccessful attempt,CW is 978-0-7695-5013-8/13 $31.00 © 2013 IEEE DOI 10.1109/ICMIRA.2013.23
RELATED WORK
87
q=10 (4) depending on the transition of the array from one status to another.
value only after fixed time interval.station calculates contention window value taking into account its local information about collisions and successful transmissions incurred.Though it reduces the energy consumption and overhead of computations ,this proposal in itself has lack of reactivity in its very definition.Bianchi in [7] uses the current contending stations to adjust the contention window value.usage of dynamic network parameter to determine back off period increases the computational overhead.In [8] the main feature of the proposal entails the fact that back off time is set using value in three element vector.The value of vector is based on both the sensing of medium(busy or idle) as well as on status of each transmission trial.To keep into account sensing of medium for CW determination hold an anamoly.This could be explained through a case that station found the medium idle,changed the back off time accordingly .After required wait time ,finding the medium idle,it transmits the frame and collision is meted by the frame.This is because actual network condition was otherwise.Therefore the use of medium sensing is not efficient for network status prediction.In our proposal value of back off time is dependent on the record of status of last two transmissions.Credebility of our proposal is high as it depends on information regarding transmitted packet in the channel. Evaluation of this network parameter for determining network condition is dependent on actual transmission results,thus proves efficient in increasing throughput of the network.
Fig 1.Contention Window Assessment Function
V.
In this section, we study the performance of the RBCW in comparison with IEEE 802.11 DCF by using the MATLAB TOOL (version 7.77) .The simulation enviorment consists of multiple sender-receiver pairs connected with bandwidth of 10Mbps.Experiments were performed to measure the performance of data transfer over wireless link with the RBCW scheme and the IEEE 802.11 standard.we generated random motion of contending stations as a function of variable with mean zero and variance one . All the other parameters deployed in the simulation model are listed in table 1.
. IV.
PROPOSED MECHANISM
Table1
Fundamental behind our algorithm defines contention window value for every transmission using information regarding status of latest transmission and the one previous to it.In the proposal ,each station uses its own record of last two transmissions to set the contention window.Each station deploys two arrays of size of 2 elements each.Array α has binary values depicting history of last two transmissions of station(‘0’ for failure and ‘1’ for success).Initially the value of contention window is set to its minimum CWmin.values in arrays are also intialized to ones. With each transmission current status of array α is stored in array β.Subsequently,oldest of transmission status is removed and the latest is added to it.for the next transmission CW is chosen using (1),where x is determined on the basis of status of both arrays.As shown in (4),value of q is carefully chosen after repeated simulations.channel conditions considered are ideal ,i.e.no hidden and exposed node problem is taken in simulation scenario.Our proposal involves extra memory space and additional operations to be performed by the station .This is compensated by the desired results achieved by RACW. CW=CWmin ± pq where, CWmin=3 p=0,1,2
SIMULATION AND ANALYSIS
Payload Packet SIFS
8184 bits 28µs
DIFS
128µs
ACK Slot time Data Rate
112 bits 50µs 10mbps )
Analysis of the proposed model was done in different scenarios. Performance metrics taken into consideration were throughput, packet delivery ratio and transmission delay. In first simulation scenario, system throughput and PDR, were measured for varied contending stations for both the mechanisms.As shown in fig 2,throughput of RACW shows better performance over DCF.RACW increases the performance of network for both small and large number of stations.Throughput was normalized with respect to the channel capacity and measured only with payload being sent as data packet.Fig 3shows better packet delivery performance of RACW than DCF in small sized network.However, we note that packet delivery ratio for network size 20 to 30 shows approximately same results for
(1) (2) (3)
88
both the mechanisms. As depicted,RACW does offer better performance,there lies a tradeoff between increased throughput and transmission delay.Increase in computational time in RACW increases the transmission delay with respect to DCF.Fig4 shows increase of transmission delay with increase in contending stations.Other point depicted in graph is the peculiar behavior of DCF under varied network size.With varied network size,transmission delay shows very little variation.
0.012 DCF RACW 0.01
Transmission Delay 0.008
0.006
0.004 -5
x 10
10
0.002
DCF RACW
9
0 8
5
10
15
20
25 30 35 Number Of Stations
40
45
50
Fig 4.Tradeoff representation.
Throughput 7
CONCLUSION 6
In this paper,we have presented a scheme for increasing the efficiency of MAC layer approach for collision avoidance,thereby increasing the performance of wireless LANs.Fundamental idea in proposal is to utilize the history of data transmissions for estimating current network conditions and ineffect optimize the contention window size accordingly. Simulation studies show the results depicting significant improvement in throughput of network as well as in PDR.For future work,we want to show the effect of noise as well of more intricate conditions on the performance of the proposed scheme.
5
4
3
5
10
15
20
25 30 35 Number Of Stations
40
45
50
Fig 2.Throughput versus contending stations
25
ACKNOWLEDGEMENT
DCF RACW
I would like to extend my gratitude to my supervisors Prabjot Kaur and Anuradha for their constant guidance and support required throughout the duration of this study.
20 Packet Delivery Ration % 15
REFERENCES 10
[1] V. Bharghavan, A. Demers, S. Shenker and L. Zhang, "MACAW: A media access protocol for wireless LAN's". In Proceedings of ACM SIGCOMM 1994, United Kingdom, pp. 212-25, 1994. [2] IEEE std 802.11,Wireless LAN Media Access Control (MAC) and Physical Layer Specifications,2012. [3] Hao Li wang,MAC layer approaches for security and performance enhancement in IEEE 802.11,thesis submitted for degree of doctorate of philosophy,2004,Iowa state university. [4] G.Bianchi, “IEEE 802.11-Saturation throughput analysis”,IEEE communi letter ,vol 2 ,pg318-320,Dec 1998. [5] Bianchi,G:Performance analysis of the IEEE 802.11 distributed Coordination function:IEEE journal on selected areas in communications 18(3) (2000)535-547 [6]Bianchi,G.Fratta,L.,Oliveri,M.:Performance evaluation and enhancement of the CSMA/CA MAC protocol for 802.11 wireless LANs In:Proc. Of PIMRC’96.(1996). [7] T. Razafindralambo and I. Gu´erin Lassous,SBA:A simple back off algorithm for wireless adhoc networks. Fratta et al. (Eds.): NETWORKING 2009, LNCS 5550, pp. 416–428, 2009. [8] Balador A,Movaghar A,Jabbehdari S,Kanellopoulos D.A novel Contention window control scheme for IEEE 802.11 WLANs.
5
0
5
10
15
20
25 30 35 Number Of Stations
40
45
50
Fig 3.PDR versus contending stations
89
