1528 ieee transactions on parallel and distributed …

Enhancing VANET Performance by JointAdaptation of Transmission Power

and Contention Window SizeDanda B. Rawat, Member, IEEE, Dimitrie C. Popescu, Senior Member, IEEE,

Gongjun Yan, Member, IEEE, and Stephan Olariu, Senior Member, IEEE

Abstract—In this paper, we present a new scheme for dynamic adaptation of transmission power and contention window (CW) size to

enhance performance of information dissemination in Vehicular Ad-hoc Networks (VANETs). The proposed scheme incorporates the

Enhanced Distributed Channel Access (EDCA) mechanism of 802.11e and uses a joint approach to adapt transmission power at the

physical (PHY) layer and quality-of-service (QoS) parameters at the medium access control (MAC) layer. In our scheme, transmission

power is adapted based on the estimated local vehicle density to change the transmission range dynamically, while the CW size is

adapted according to the instantaneous collision rate to enable service differentiation. In the interest of promoting timely propagation of

information, VANET advisories are prioritized according to their urgency and the EDCA mechanism is employed for their dissemination.

The performance of the proposed joint adaptation scheme was evaluated using the ns-2 simulator with added EDCA support. Extensive

simulations have demonstrated that our scheme features significantly better throughput and lower average end-to-end delay compared

with a similar scheme with static parameters.

Index Terms—Vehicular networks, VANETs, broadcast, contention window adaptation, message differentiation, transmission power

adaptation, QoS, medium access control protocol, 802.11e EDCA, intelligent transportation system.

Ç

1 INTRODUCTION

WE are witnessing an unmistakable convergence ofVehicular Ad-hoc Networks (VANETs) and Intelligent

Transportation Systems (ITSs) that is poised to bring about arevolutionary leap by making our roadways and streets saferand the driving experience more enjoyable [1]. Working intandem with the fielded ITS infrastructure, VANET isexpected to enhance the awareness of the traveling publicby aggregating, propagating, and disseminating up-to-the-minute information about existing or impending traffic-related events. In support of their mission, VANET commu-nications, employing a combination of Vehicle-to-Vehicle(V2V) and Vehicle-to-Infrastructure (V2I) wireless commu-nication are expected to integrate the driving experience intoa ubiquitous and pervasive network that will enable noveltraffic monitoring and incident detection paradigms [2], [3]. Itis worth noting that the vast majority of the traffic advisoriesare of a general interest and, therefore, benefit from beingbroadcast. For instance, when a traffic incident occurs, all thevehicles on the road benefit from timely and accurate

information dissemination allowing the drivers to makeinformed decisions. Thus, reliability and low delay areextremely important factors in VANET safety applications.It is widely known that, due to high-speed mobility, V2V andV2I communication links tend to be shortlived. Thus, it isimportant to propagate traffic-related information toward acertain region of interest instead of sending to a particularvehicle; moreover, one of the best ways of propagating traffic-related advisories toward a particular region is some form of(controlled) broadcast transmission.

One strategy of increasing duration of communicationlinks in VANET is by increasing the transmission range insparse traffic conditions, where only a few vehicles may bepresent on the road. However, increasing the transmissionrange may generate high levels of disruptive interferenceand high-network overhead in dense traffic conditions. Itfollows that dynamic adaptation of transmission power inresponse to changing traffic density is a critical requirementin VANET. In addition, in order to propagate emergencymessages in a timely manner, VANET must support someform of message differentiation, similar in spirit to servicedifferentiation for QoS in the contention-based channelaccess mechanism EDCA of 802.11e [4]. To implement thisstrategy, different priority levels can be assigned to varioustraffic-related messages according to their urgency or delayrequirements. For example, messages related to an incidenton the roadway should be propagated to the target regionon time and in an accurate manner in order to avoidcongestion and potential secondary accidents.

1.1 Related Work

Because of the advantages it offers, the IEEE 802.11 wirelessstandard is used by a host of protocols for information

1528 IEEE TRANSACTIONS ON PARALLEL AND DISTRIBUTED SYSTEMS, VOL. 22, NO. 9, SEPTEMBER 2011

. D.B. Rawat and D.C. Popescu are with the Department of Electrical andComputer Engineering, Old Dominion University, 231 Kaufman Hall,Norfolk, VA 23529. E-mail: {drawa001, dpopescu}@odu.edu.

. G. Yan is with the Department of Natural, Information, and MathematicalSciences, Indiana University, Kokomo, IN 46904-9003.E-mail: [email protected].

. S. Olariu is with the Department of Computer Science, Old DominionUniversity, Norfolk, VA 23529. E-mail: [email protected].

Manuscript received 11 May 2010; revised 30 Sept. 2010; accepted 5 Oct.2010; published online 18 Jan. 2011.Recommended for acceptance by A. Nayak.For information on obtaining reprints of this article, please send e-mail to:[email protected], and reference IEEECS Log Number TPDS-2010-05-0290.Digital Object Identifier no. 10.1109/TPDS.2011.41.

1045-9219/11/$26.00 � 2011 IEEE Published by the IEEE Computer Society

forwarding and dissemination in VANET. In 2004, the IEEETask Group p (TGp) started to develop IEEE 802.11p [5] byamending the IEEE 802.11 standard to include vehicularcommunication operating at speeds up to 200 km/h andwith a communication range as large as 1,000 m. The IEEE802.11p standard was designed to operate at 5.9 GHz anddata rates up to 27 Mb/s with the stated goal of supportingV2I communication or V2V communication in the context ofthe FCC-mandated Dedicated Short Range Communication(DSRC) [6].

To set the stage for the proposed scheme, it is appropriateto note that the use of the EDCA mechanism of 802.11e in thecontext of VANET was discussed by Suthaputchakun andGanz [7], where a priority-based scheme for V2V commu-nications was proposed. However, while the authors of [7]proposed incorporating of the EDCA mechanism of 802.11ein VANET, they did not address the problem of adaptingQoS parameters or that of adapting transmission poweraccording to local traffic conditions.

To the best of our knowledge the problem of adaptingtransmission power in VANET based on vehicle densitywas first discussed by Artimy [8], while the problem ofdynamically adapting the CW size for reliable broadcast inVANET was discussed by Balon and Guo [9]. However, theauthors of [9] only considered the channel access timeaccording to the urgency of messages and their delayrequirements, without considering the adaptation of trans-mission power, or the prioritization of messages accordingto their urgency, or the adaptation of the CW size fortransmission opportunity, which can enhance systemthroughput while reducing end-to-end message delay[10], [11].

1.2 Our Contribution

Our work was motivated by the observation that theexisting schemes [8], [9] did not take into account thedynamically changing topology of VANET and, conse-quently, kept either the transmission power of a vehicle orthe QoS related parameters fixed. Indeed, the majorcontribution of this work is a new scheme for jointadaptation of transmission power at the PHY layer and ofthe CW size at the MAC layer, according to local vehicledensity and network condition, respectively.

The proposed scheme adapts transmission power dyna-mically based on estimated local traffic/vehicle density. Forestimating local vehicle density, we use a differentapproach than the one proposed by Artimy [8]; as it turnsout, our traffic density approximation is more accurate thanthe one in [8], resulting in a more appropriate transmissionrange. In addition, we prioritize messages according to theirurgency by incorporating IEEE 802.11e EDCA for timelypropagation of high-priority messages toward the destina-tion region.

We assume that the vehicles start running the proposedscheme as soon as they receive traffic-related messages andnote that issues related to incident detection, admissioncontrol, as well as security and privacy of transmittedmessages and participating vehicles, which are importantfor ITS, are outside of the scope of this paper.

The remainder of this work is organized as follows: InSection 2, we discuss the EDCA mechanism of 802.11e and

its role in VANET, and we formally state our problem. InSection 3, we present the proposed scheme for jointdynamic adaptation of transmission power and CW valuesbased on vehicle density estimation and network condition.In Section 4, we formally state the algorithm for joint powerand CW adaptation followed by presentation of numericalresults obtained from simulations in Section 5. Finally,Section 6 offers concluding remarks and directions forfuture investigations.

2 THE EDCA MECHANISM OF IEEE 802.11e AND

ITS ROLE IN VANET

The IEEE 802.11 standard plays a major role in wirelessnetworking. Due to their simplicity, scalability, flexibility,and cost effectiveness, wireless local area networks(WLANs) based on IEEE 802.11 are among the most widelydeployed WLAN technologies. The fundamental accessmechanism of IEEE 802.11 is applicable to VANETcommunications, which use IEEE 802.11p [5], a modifiedversion of IEEE 802.11a.

It is widely known that the baseline IEEE 802.11 standarddoes not provide for the service differentiation necessary forsupporting QoS for time critical data such as voice traffic inWLAN [12], [13]. In order to address the issue of servicedifferentiation the IEEE 802.11e standard [4] specifies thedistributed contention-based channel access mechanism,referred to as EDCA. The EDCA is available in the ad hocmode, where no infrastructure is available. The EDCAscheme relies on CSMA/CA along with a slotted BinaryExponential Backoff (BEB) mechanism for contention-basedchannel access [4] and supports MAC-level QoS andprioritization of different data/traffic by defining multipleAccess Categories (ACs) with different CW and ArbitrationInterframe Space (AIFS) values.

According to [4] a station with QoS implements fouraccess categories and there is a set of EDCA parametersassociated with each AC. These parameters include AIFS[AC] and CW with its minimum and maximum valueCWmin [AC] and CWmax [AC], respectively. Each ACfrom every station independently starts a backoff timer afterdetecting that the channel is idle for an AIFS [AC] intervaland competes with other ACs for channel access and theopportunity to transmit. For each AC, the backoff period isselected from a uniform distribution over [0, CW [AC]]. TheCW size is initially assigned CWmin and doubles whentransmission fails, up to CWmax. The CW size is reinitia-lized when the CW reached CWmax, and the process isrepeated. We note that the smaller CWmin [AC], the shorterthe channel access delay for the corresponding priority, andhence the better the chance of the station to access thechannel in a given traffic condition. When an application isadmitted, it has a number of QoS parameters. If two or morebackoff timers within the same station finish backoff at thesame time, there will be a virtual collision which will besolved by the station’s internal scheduler.

We note that MAC protocols for VANET have to considerdifferent types of traffic messages as well as a rapidlychanging network topology. For example, it is highlydesirable for emergency messages related to traffic incidentson the roadway to have higher priority than other messages

RAWAT ET AL.: ENHANCING VANET PERFORMANCE BY JOINT ADAPTATION OF TRANSMISSION POWER AND CONTENTION WINDOW... 1529

in order to get rapid channel access, and thus prioritization ofdifferent messages according to their urgency is an im-portant requirement in VANET. As a consequence, incorpor-ating EDCA in VANET enables better messagedifferentiation and ensures that the high-priority messagesget transmission opportunities on a preferential basis.

Furthermore, in congested traffic the network topology isvery dense, while, as soon as the congested region is passedthe network topology may become sparse again. In thecontext of broadcast transmission used for V2V commu-nication, high-transmission power in a region with high-vehicle density results in high-network load. Therefore,vehicle density in a given region is a useful metric foradapting the transmission power.

We present a new method which combines the advan-tage of dynamic adaptation of transmission power at thePHY layer as a function of vehicle density with dynamicadaptation of CW size in EDCA at the MAC layer toenhance the performance of V2V communications inVANET. The proposed approach ensures that propagationand dissemination of prioritized messages will occur withhigh throughput and low end-to-end delay.

3 TRANSMISSION POWER ADAPTATION AND

PRIORITIZATION OF MESSAGES

In this section, we describe how transmission range andtransmission power are calculated based on local density ofvehicles and network conditions, and how different mes-sages are assigned different priorities based on their urgency.

One of the starting points of our investigation wasprovided by the following expression derived in Artimy [8]for the transmission range (TR) based on the estimated localvehicle density

TR ¼ min

�Lð1�KÞ;

ffiffiffiffiffiffiffiffiffiffiffiffiffiL lnL

K

rþ �L

�; ð1Þ

where

. � is a constant from traffic flow theory [8],

. L is the length of the road segment over which thevehicle estimates its initial local vehicle density, and

. K is the local vehicle density for a given vehicle,calculated as the ratio K ¼ AN

TN of the actual number(AN) of vehicles on the road that are present withinits transmission range to the total number (TN) ofvehicles that can be present on the road for currenttransmission range, travel speed and safety separa-tion distance as shown schematically in Fig. 1.

We note that the method used for estimating K in [8] isbased solely on the vehicle’s movement and may not alwaysgive a good estimate of the local traffic density K. Forinstance, when a given vehicle moves at low speed themethod in [8] will estimate that the local vehicle density K

is high, while when it moves at high speed it will estimate alow vehicle density K. In order to improve the accuracy ofthe local vehicle density estimate K, we employ a differentmethod which uses information obtained from the interac-tion of a given vehicle with other vehicles in the network.For this purpose, we note that in the DSRC standard a givenvehicle exchanges its status with neighboring vehiclesapproximately 10 times each second [14], and individualvehicles can use this information to estimate the actualnumber of vehicles AN in their vicinity by using the 12-bitsequence number of the IEEE 802.11 MAC header. We notealso note that this method does not introduce significantnetwork overhead since it exploits the periodic message inDSRC enabled systems, and that a similar approach hasbeen used in [9] in a different context where it was used tomeasure the collision.

For example, consider that the current TR for a givenvehicle is 600 m that is obtained from (1) with L ¼ 1;000 m

and the vehicles on the same lane maintain average safetyseparation distance of 20 m. For a separated highway withtwo lanes of traffic in each direction, the TN for the givenvehicle is calculated as TN � ½600=20� � 2 � 2 ¼ 120, whilethe AN for the given vehicle is calculated based on thereceived information from its neighboring vehicles. As-suming, for example, that AN ¼ 65 we calculate theestimated vehicle density K ¼ AN=TN � 0:54. Using thisvalue of K the given vehicle will update its transmissionrange using (1).


Fig. 1. Estimating the local vehicle density K on the road for a given vehicle for its transmission range (TR). (a) TN of possible reachable neighborvehicles. (b) AN of reachable neighbor vehicles.

The transmission range TR obtained by using the

outlined method for estimating K was compared to that

in [8] and the results, plotted in Fig. 2,1 show that our

method provides a better estimates of the local vehicle

density and, consequently, of the normalized transmission

range TRL , which turns out to be closer to the theoretical

value obtained when the actual vehicle density K is used in

(1) instead of its estimates.

3.1 Transmission Power Adaptation

Once the transmission range is obtained using (1), we need

to map it to an actual transmission power value. For this

purpose, we use the lookup table captured in Table 1

containing the transmit power values corresponding to

different transmission ranges. We note that the data in

Table 1 was obtained by simulations of basic wireless

propagation models for different VANET scenarios and aspecific power value is assigned for a given transmissionrange interval to include urban, city, and rural environ-ments. We also note that the lookup table approach is fastersince no computations are required.

3.2 Prioritization of Messages

As discussed in Section 2, the IEEE 802.11e EDCA has theservice differentiation to provide QoS for different types ofmessages: voice traffic, video traffic, best effort traffic, andbackground traffic [4]. To incorporate the EDCA mechan-ism in VANET, we categorize the different messagesaccording to their urgency and delay requirements [7] aslisted in Table 2.

The different access categories in EDCA will havedifferent QoS parameters associated with them. Table 3gives the QoS parameters corresponding to the ACs in802.11e EDCA. The higher the access category number, thehigher the channel access or transmission opportunity willbe. That means the CWmin value for AC(3) will be the leastamong all ACs. The backoff counter drawn uniformly from[0, CW [AC]] will have an initial value of CWmin, implyingthat AC(3) will get the highest transmission opportunityover others. Moreover, high priority classes in turn use ashorter interframe spacing (IFS) and a smaller CW size, sothat they will get preferential treatment over lower priorityclasses. Each vehicle will have four different queues, one foreach priority class with a virtual collision handler to handleinternal collisions.

3.3 Contention Window Size Adaptation

In order to support message differentiation for differenttypes of messages listed in Table 2 the size of the CW inTable 3 should also be adapted taking into account the factthat vehicles that have higher priority messages should notget the chance to be greedy (i.e., get channel access most ofthe time) while higher priority messages should not bewaiting for a long time for the opportunity to transmit.

We note that a vehicle attempting to get transmissionopportunity must wait for the channel to remain idle for theduration of the AIFS before starting its backoff timer. Wealso note that holding channel access for a long time forhigher priority messages may result in a delay in messagepropagation which will not be able to notify and/or preventincidents on the roadway, such as congestion and traffic-jam buildups.


Fig. 2. Comparison of normalized transmission range (TR=L) versuslocal vehicle density K for the proposed scheme (that is running onlyStage 1 of Algorithm 1) and for the method used in [8].

TABLE 1Lookup Table for Transmission Power Corresponding to a

Given Transmission Range

TABLE 2Message Priorities [7]

TABLE 3Priority Specific Parameters [4]

1. The simulation setup used in obtaining Fig. 2 is described in thesupplementary electronic document file, which can be found on theComputer Society Digital Library at http://doi.ieeecomputersociety.org/10.1109/TPDS.2011.41.

Therefore, to mitigate these adverse effects the dynamicadaptation of QoS parameters, in particular CW sizes, fordifferent access categories is essential since the backoffcounter value is obtained uniformly from [0, CW [AC]] andthe initial CW value is CWmin. In our proposed approach,the size of the CW may either increase or decrease, and CWadaptation is carried out by applying the well-knownapproach used in the IEEE 802.11 algorithms by which thesize of window CW [AC] is varied by a factor of two. Inother words, the window size is doubled if one has toincrease the size, and is reduced by half if one has todecrease its size. The CW size will continue to increase untilit reaches to maximum size of the window, CWmax [AC]after which it will be reinitialized to CWmin. Thus, in ourproposed approach the window size fluctuates according tothe network conditions observed by a vehicle while inconventional 802.11 technologies (including EDCA) the sizeof the window remains fixed no matter what the networkcondition is. We note that the increase in CW [AC] values(for all ACs maintaining the hierarchy of CW [AC] values asin EDCA [4]) when the network is congested, will give lessopportunity for all ACs to reduce network load because ofbroadcast and rebroadcast. Similarly the decrease in CW[AC] values (for all ACs maintaining hierarchy of CWvalues) when the network has less or no collision, will givehigher opportunity for all ACs. In both cases, the prefer-ential treatment is preserved by hierarchical increments incorresponding CW [AC] values.

The local state of the network can be determined as in [9]by using the record of sequence numbers corresponding toindividual vehicles from which it receives messages. Byusing lost sequence numbers each vehicle can calculate theapproximate percentage of lost frames sent by othervehicles to in a given period of time. Using these statisticsand based on the local reception rate, each vehicledetermines the local state of the network as suggested byBalon and Guo [9] and can use it to adapt the CWparameters in the EDCA mechanism. The adaptation ofCW size according to network conditions results in highthroughput and lower delay for high-priority messages,while lower priority messages also get channel access butwith lower preference over higher priority ones [10].

The throughput obtained by the proposed scheme usingdynamic adaptation of CW values based on networkcondition was compared to the one reported by Chen et al.in [11] and the results, plotted in Fig. 3,2 show that ourscheme outperforms the one in [11].

4 THE ALGORITHM

Based on the methods discussed in the previous section, wepresent an algorithm which adapts both transmitted powerand CW size and which should be run by individualvehicles periodically to ensure that proper updates oftransmission power and CW [AC] values occur accordingto the local vehicle density and the network condition,respectively. The algorithm is formally stated as Algorithm 1

and consists of two stages: transmit power adaptation andCW size adaptation.

4.1 Dynamic Adaptation of Transmission Power

Initially, individual vehicles start with an arbitrary trans-mission power and listen for information from othervehicles. Once a vehicle receives message packets fromother vehicles, it starts to analyze the sequence numbersand to count the vehicles around its locale. In order tomitigate the adverse effects of high-transmission power andto increase the duration of the communication link in caseof low traffic density for intervehicle communication, eachvehicle dynamically adapts its transmission power based onthe estimated local vehicle density. The vehicle densitywithin its transmission range is calculated using the methoddiscussed in Section 3.1 which is based on observation ofpackets that are currently received and does not introducesignificant network overhead to identify the neighbors ofthe vehicle [9]. Using the estimated vehicle density thealgorithm calculates the transmission range using (1), andthen sets up the corresponding transmission power usingthe look up Table 1.

We note that maximum transmit power corresponding tomaximum transmission range is selected when either thelocal vehicle density K is sparse, that is, lower than someapplication-dependent threshold value �1 or when thevehicle needs to transmit priority 1 messages.3 The choiceof threshold value �1 plays a significant role in theimplementation of the algorithm and is also adaptedaccording to the local density observed by a given vehicleperiodically based on the threshold history for the givenvehicle. Moreover, since the algorithm is employed in adistributed manner threshold values for different vehiclesmight be different.


Fig. 3. Throughput versus simulation time for proposed scheme (i.e., bykeeping transmission power fixed and without running the poweradaptation steps of Algorithm 1) and the method used in Chen et al.[11] for higher priority messages (AC(2) and AC(3)).

2. The simulation setup used to obtain Fig. 3 is described in thesupplementary electronic document file, which can be found on theComputer Society Digital Library at http://doi.ieeecomputersociety.org/10.1109/TPDS.2011.41.

3. Priority 1 messages are transmitted in case of accidents or fromemergency vehicles as shown Table 3. They have low delay requirementsand should propagate on time in a single hop (if possible) within themaximum transmission range as all vehicles in the destination region seekemergency related messages, so that they respond to the situation accordingto the message received.

Algorithm 1. Joint Adaptation of Transmit Power andCW Size.

4.2 Dynamic Adaptation of CW Size

Dynamic adaptation of CW size causes changes in thebackoff counter so that timely transmission of messagesoccurs according to the network conditions, U, theperceived collision rate and local vehicle density. The CWsize adaptation is performed as discussed in Section 3.3, inresponse to network conditions estimated by analyzing thereceived sequence numbers at MAC layer as discussed inSection 3. The estimated collision rate is an indication ofhow congested the network is and how information flowfrom a vehicle should be controlled. The dynamic adapta-tion of the CW size is regulated by a threshold value �2. Wenote that, as it was the case with the power adaptation,proper choice of the threshold value �2 in CW adaptation isimportant and affects the performance of the system, andthat the threshold �2 may also be adapted periodically based

on the network conditions and on the threshold history for agiven vehicle.

5 SIMULATIONS AND NUMERICAL RESULTS

In order to illustrate the performance of the proposed jointdynamic adaptation scheme, we have simulated Algorithm1 and have compared it with the default EDCA schemewhich has fixed values for transmission power and QoSparameters [4]. The simulation setup is described in thesupplementary electronic document file, which can befound on the Computer Society Digital Library at http://doi.ieeecomputersociety.org/10.1109/TPDS.2011.41 and itssimulation parameters are summarized in Table 4.

In order to incorporate the EDCA mechanism in VANETusing ns-2, we have mapped the suitable messages with thecorresponding service differentiated EDCA access cate-gories as listed in Table 2, and have assigned the QoSparameters for each AC once the message becomesavailable at the vehicle.

We illustrate the performance of the proposed jointdynamic adaptation scheme for transmit power and conten-tion window by comparing it with the default scheme in termsof the overall throughput and end-to-end message delay in ascenario consisting of approximately 50 percent highestpriority messages AC (3) and approximately 50 percent othertypes of messages AC (0)-AC (2) in the network. Thiscorresponds to the third simulation experiment as describedin the supplementary electronic document file, which can befound on the Computer Society Digital Library at http://doi.ieeecomputersociety.org/10.1109/TPDS.2011.41. The


TABLE 4Simulation Parameters

Fig. 4. Overall throughput variation.

throughput variation is plotted Fig. 4, and the average end-to-end message delay is plotted in Fig. 5.

From Fig. 4, we note that initially (up to approximately48 s) the throughput value is similar for both schemes sincethe vehicles might not be able at first to detect theirneighbors, in order to estimate vehicle density and adjusttheir transmission power and CW size values. As soon asAlgorithm 1 start adjusting the transmission power and CWsize according to vehicle density and network condition, theoverall throughput of the proposed scheme is higher thanthat of the default scheme.

Fig. 5 shows the average end-to-end delay for 1) allmessages, 2) the highest priority—AC(3) only—messages,and 3) the other AC(0)—AC(2) categories (all message typesexcept highest priority). In all cases there is no delay in the

beginning stage (up to approximately 48 s) since all messagesget transmission opportunity as soon as they are available atthe vehicle. Delay at around 50 s is high since vehicles couldnot be able to adapt the power and CW values according to thevehicle density and network condition during the initial stageof the simulation.

As simulation time increases, the delay decreases since

individual vehicles adapt their transmission power and CW

values dynamically according to vehicle density and net-

work condition. As can be observed from Fig. 5 the average

end-to-end delay is lower for the proposed scheme than for

the default scheme in all cases.

6 CONCLUSION

In this paper, we presented a new scheme for reliable

broadcast transmission in vehicular communication with

joint dynamic adaptation of transmission power and CW

size. The scheme incorporates the EDCA medium access

mechanism of IEEE 802.11e in VANET to set priority for

different messages according to their urgency, and consists

of an algorithm by which individual vehicles dynamically

adapt transmission powers according to the estimated local

vehicle densities and adjust CW [AC] for all ACs based on

data collision rate on the network.Performance of the proposed scheme is illustrated with

numerical results obtained from simulations which show

that better throughput is achieved with lower delay than

when the default scheme is used.

ACKNOWLEDGMENTS

The authors are grateful to the anonymous reviewers fortheir constructive comments on the paper. This work wassupported in part by the National Science Foundation(NSF) under grant CNS-0721586 and was presented inpart at the 70th IEEE Vehicular Technology Conference—VTC 2009 Fall.

REFERENCES

[1] W. Chen, S. Cai, and T. Inc, “Ad Hoc Peer-To-Peer NetworkArchitecture for Vehicle Safety Communications,” IEEE Comm.Magazine, vol. 43, no. 4, pp. 100-107, Apr. 2005.

[2] O. Riva, T. Nadeem, C. Borcea, and L. Iftode, “Context-AwareMigratory Services in Ad-Hoc Networks,” IEEE Trans. MobileComputing, vol. 6, no. 12, pp. 1313-1328, Dec. 2007.

[3] J. Nzounta, N. Rajgure, G. Wang, and C. Borcea, “VANET Routingon City Roads using Real-Time Vehicular Traffic Information,”IEEE Trans. Vehicular Technology, vol. 58, no. 7, pp. 3609-3626, Sept.2009.

[4] Part 11: Wireless LAN Medium Access Control (MAC) andPhysical Layer (PHY) Specifications Amendment 8: MediumAccess Control (MAC) Quality of Service Enhancements, IEEEStd. 802.11e, 2005.

[5] D. Jiang and L. Delgrossi, “IEEE 802.11p: Towards an Interna-tional Standard for Wireless Access in Vehicular Environments,”Proc. IEEE Vehicular Technology Conf. (VTC ’08-Spring), pp. 2036-2040, May 2008.

[6] R. Sengupta and Q. Xu, “DSRC for Safety Systems,” CaliforniaPATH-Partners for Advanced Transit and Highways, vol. 10,no. 4, pp. 2-5, 2004.

[7] C. Suthaputchakun and A. Ganz, “Priority Based Inter-VehicleCommunication in Vehicular Ad-Hoc Networks using IEEE802.11e,” Proc. IEEE 65th Vehicular Technologies Conf. (VTC ’07-Spring), pp. 2595-2599, Apr. 2007.


Fig. 5. Average end-to-end message delay. (a) For all messages. (b) ForAC(0), AC(1), and AC(2). (c) For AC(3).

[8] M.M. Artimy, “Assignment of Dynamic Transmission RangeBased on Estimation of Vehicle Density,” IEEE Trans. IntelligentTransportation Systems, vol. 8, no. 3, pp. 400-412, Sept. 2007.

[9] N. Balon and J. Guo, “Increasing Broadcast Reliability inVehicular Ad Hoc Networks,” Proc. Third ACM Int’l WorkshopVehicular Ad Hoc Networks (VANET ’06), pp. 104-105, Sept. 2006.

[10] D.B. Rawat, D.C. Popescu, and M. Song, “Performance Enhance-ment of EDCA Access Mechanism of IEEE 802.11e WirelessLAN,” Proc. IEEE Radio and Wireless Symp. (RWS ’08), pp. 507-510,Jan. 2008.

[11] X. Chen, H. Zhai, X. Tian, and Y. Fang, “Supporting QoS in IEEE802.11e Wireless LANs,” IEEE Trans. Wireless Comm., vol. 5, no. 8,pp. 2217-2227, Aug. 2006.

[12] G.-M. Muntean and N. Cranley, “Resource Efficient Quality-Oriented Wireless Broadcasting of Adaptive Multimedia Con-tent,” IEEE Trans. Broadcasting, vol. 53, no. 1, pp. 362-368, Mar.2007.

[13] Y. Xiao, “Performance Analysis of Priority Schemes for IEEE802.11 and IEEE 802.11e Wireless LANs,” IEEE Trans. WirelessComm., vol. 4, no. 4, pp. 1506-1515, Aug. 2005.

[14] “Vehicle Safety Communications Project Task 3 Final Report:Identify Intelligent Vehicle Safety Applications Enabled byDSRC,” Vehicle Safety Communications Consortium Consistingof, BMW, DaimlerChrysler, Ford, GM, Nissian, Toyota, and VW.,Mar. 2005.

Danda B. Rawat received the bachelor’s degreein computer engineering in 2002, the master’sdegree in information and communication en-gineering from the Tribhuvan University, Kath-mandu, Nepal, in 2005, and the PhD degree inelectrical and computer engineering from OldDominion University, in 2010. His researchinterests include areas of wireless communica-tions and wireless cellular/ad-hoc networks. Heis a member of the IEEE.

Dimitrie C. Popescu received the engineeringdiploma and MS degrees from the PolytechnicInstitute of Bucharest, Romania, and the PhDdegree from Rutgers University, all in electricalengineering. He is currently an assistant pro-fessor in the Department of Electrical andComputer Engineering, Old Dominion Univer-sity. His research interests include areas ofwireless communications, digital signal proces-sing, and control theory. He is an associate

editor for IEEE Communications Letters, he has served as technicalprogram chair for the vehicular communications track of the IEEE VTC2009 Fall, finance chair for the IEEE MSC 2008, and technical programcommittee member for the IEEE GLOBECOM, ICC, WCNC, and VTCconferences. He is a senior member of the IEEE.

Gongjun Yan received the PhD degree incomputer science from Old Dominion Univer-sity, in 2010, and is currently an assistantprofessor in the Department of Natural, Infor-mation, and Mathematical Sciences, IndianaUniversity Kokomo. He has been working on theissues surrounding Vehicular Ad-Hoc Networks,Sensor Networks, and Wireless Communica-tion. His research interests include security,privacy, routing, and healthcare. He applies

mathematical analysis to model behavior of complex systems andintegrates existing techniques to provide comprehensive solutions. Heis a member of the IEEE.

Stephan Olariu received the PhD degree incomputer science from the McGill University, in1986. He is currently a professor in theComputer Science Department, Old DominionUniversity, and is a world-renowned technologistin the areas of wireless networks, mobile multi-media systems, parallel and distributed systems,and architectures and networks. He was invitedand visited more than 120 universities andresearch institutes around the world lecturing

on topics ranging from wireless networks and mobile computing, tobiology-inspired algorithms and applications, to telemedicine, to wirelesslocation systems, and security. He is an associate editor of Networksand IEEE Transactions on Parallel and Distributed Systems and serveson the editorial board of Journal of Parallel and Distributed Computing,Journal of Ad hoc and Sensor Networks, and Parallel, Emergent andDistributed Systems. He is a senior member of the IEEE.

. For more information on this or any other computing topic,please visit our Digital Library at www.computer.org/publications/dlib.


1528 ieee transactions on parallel and distributed …

Documents