Impact of Two Realistic Mobility Model for Vehicular Safety Applications

Transcription

1 Impact of Two Realistic Model fo Vehicula Safety Applications Md. Habibu Rahman Dept. of Compute Science Ameican Intenational Univesity-Bangladesh Dhaka, Bangladesh Mohammad Nasiuddin Dept. of Compute Science Univ. Genoble Alpes Genoble, Fance Abstact Vehicula safety applications intended fo VANETs. It can be sepaated by inte-vehicle communication. It is needed fo a vehicle can tavel safety with high velocity and must inteconnect quickly dependably. In this wok, examined the impact of the IDM-IM and IDM-LC mobility on AODV, AOMDV, DSDV and OLSR outing potocol using Nakagami popagation and IEEE p MAC potocol in a paticula uban scenaio of Dhaka city. The peiodic boadcast (PBC) agent is employed to tansmit messages between vehicles in case of emegency o collision avoidance fo vehicula safety communication. The simulation esults ecommend numeous concens such as lowe packet dop ate, delay, jitte, oute and mean-hop is necessay to be measued befoe developing a obust safety application of VANET. Keywods-VANET; AODV; AOMDV, DSDV; OLSR; IDM-IM; IDM-LC; PBC; IEEE p; Nakagami I. INTRODUCTION Vehicula Ad-Hoc Netwok (VANET) spots vehicles to communicate with each othe via vehicle-to-infastuctue (V2I) communications mode and vehicles can inteconnect with each othe via vehicle-to-vehicle (V2V) communications mode [1]. It offes the timely infomation to a dive that enables them to pedict vehicle collision avoidance, impove congestion, lane change and constuction site waning [2]. Vehicula communications systems have dawn fast eseach attention due to the potential to impove efficiency and safety of ca taffic as well as soul and popety potection. The IEEE p standad is used fo independent vehicles appoaching an intesection use Dedicated Shot Range Communications (DSRC) and Wieless Access in a Vehicula Envionment (WAVE) to peiodically send infomation such as location, heading and intesection cossing puposes to othe vehicles [3]. The signal popagation within DSRC channel can be affected by fading in an uban intesection. To sumount this touble, vaious eseach studies have demonstated that the nakagami adio popagation is athe suited fo uban scenaio [2]-[4]. In VANETs, the outing potocol opeation is egaded by vaious elements such as communication mode, vehicle/node density fluctuations, and guest/vehicle mobility patten. Fo ealistic vehicula mobility patten, VanetMobiSim can povide an actual mobility scenaio of a specific aea [5]. In this wok, the peiodic boadcast (PBC) agent, is utilized fo sending safety message between vehicles in case of emegency o collision avoidance. The IEEE p MAC potocol and Nakagami adio popagation is used to get bette pefomance on VANETs. The VanetMobiSim is used to geneate ealistic vehicula taffic of a paticula aea of Dhaka city. Fo expeimental evaluations, compaed the impact of IDM-IM and IDM-LC on AODV, AOMDV, DSDV and OLSR with espect to seveal QoS metics such as Delay, Jitte, Thoughput, and othe pefomance metics such as Aveage, Nomalized Routing Load (), Cost, and Packet Delivey Ratio (). II. RELATED RESEARCH In [2], the authos investigated that Nakagami popagation outpefoms than TwoRayGound in an uban scenaio fo AODV and OLSR outing potocol using IEEE p MAC potocol. In [4], the authos evaluated the pefomance of AODV and OLSR outing potocol using Nakagami adio popagation with IEEE p MAC potocol. In [5], the autho evaluated the pefomance of VANETs by integating clusteing of diffeent aeas and taffic lights via IDM-IM uses AODV and AOMDV potocols with two diffeent types of taffic souce connection with existing IEEE b MAC potocol and TwoRayGound popagation. In [1], the authos demonstated the impact of mobility of IEEE p pefomance of vehicula netwok by investigating cetain mobility factos such as elative speed. It holds an impotant impact on channel access at the MAC laye, bushing off the numbe of communicating nodes. In [11], the authos have offeed that p gives effective sevice diffeentiation mechanism that can be appopiate fo the mission-citical ITS application. They evaluated the MAC laye pefomance without putting on any ealistic vehicula mobility. The opeation of IEEE p MAC sub-laye, pincipally fo the V2V [1], [11], [12]. Vaious eseaches have used TwoRayGound instead of Nakagami popagation fo compaing the VANETs pefomance using diffeent outing potocol, taffic design and vaious mobility s [5]-[12]. The main dawback of thei study is ignoing the safety message tansmission in case of emegency o collision avoidance fo vehicula communications. A. Review of Routing Potocols Ad-Hoc On-Demand Distance Vecto (AODV) outing potocol enables dynamic, on-demand, self-stating, multi-hop

2 outing between paticipating mobile nodes wishing to build and sustain an ad-hoc netwok [2], [13], [14]. Ad-Hoc On- Demand Multipath Distance Vecto (AOMDV) calculates multiple loop-fee and link-disjoint paths, but clients ae unawae of the elative movement and positioning [15], [16]. Destination Sequenced Distance Vecto (DSDV) is a poactive outing potocol, whee each node maintains outing infomation fo each possible destination and can figue out the looping poblem and to cope dynamically with netwok modifications [17]. To optimize the pefomance of Optimized Link State Routing (OLSR) potocol, Multipoint Relay (MPR) nodes ae used fo the numbe of boadcasted on the netwok is minimized [18]. B. Popagation Model The Nakagami adio popagation is a mathematical geneal ing of a adio channel with fading and can povide moe configuable paametes to pemit a moe faithful epesentation of the wieless communication channel [2]. It can efficiently the chaacteistics of diffeent eal wold scenaios than TwoRayGound and capable of vaious scenaios fom fee space to modeate obstacles to high obstacles can be simulated [4]. C. Model The VanetMobiSim includes vehicle to infastuctue (V2I) and vehicle to vehicle (V2V) association [5]. It is combined the stop signals, taffic lights and activity based maco-mobility with the funding of human mobility dynamics. IDM-IM suppot smat intesection management as well as slow down and stop at intesections, o act accoding to taffic lights, if pesent [19]. In both situations, it only behaves on the fist vehicle on each oad, as IDM automatically adjusts the behavio of cas behind the leading ca. IDM-LC mobility povides oppotunities fo vehicles to alteation lane and ovetake among vehicles in the pesence of multi-lane oads. These two mattes ae high by the state of diffeent lanes such as the pating of taffic flows on dissimila lanes of the simila oad and the ovetaking itself [19]. III. SYSTEM MODEL Fig. 1 demonstated a ealistic vehicula mobility in a paticula aea of Dhaka city using IDM-IM and IDM-LC mobility to undestand the taffic condition of this aea. In IDM-IM and IDM-LC, the quantity of inteaction with a taffic light is 500. The Taffic light length is 10 seconds. The numbe of lanes is 2. The maximum numbe of multi-lane oads ae 10. Vehicle length is 5 m. The maximal acceleation of vehicle movement is 0.6 m/s 2. The comfotable deceleation of vehicle movement is 0.9 m/s 2. The minimal distance to a standing node (jam distance) is 1 m. The node s safe time headway is 0.5 s. The step fo ecalculating movement paametes is 1 s. The maximum stay duation at destination is 6 s. The minimum stay duation at destination is 2 s. The visibility distance of 200 m. The intesections located at the bodes of the map is ignoed. In IDM-LC, The politeness facto of dives when changing lane is 0.5. The theshold acceleation fo lane change is 0.5 m/s 2. Figue 1. Repesentation of Vehicula mobility fo simulation A. Simulation paamete & Quality evaluation All nodes use p MAC opeating at 6Mbps. The tansmission ange is 250 m. Fo expeimental puposes, the simulation aea is 1000 X 1000 m 2. The eal wold simulation aea is highe than the simulation aea used this study. The inteface queue length is 50 at each node. The antenna type is Omni-Antenna. TABLE I. Paamete MAC Type Channel Type Model SIMULATION PARAMETERS IEEE 80211p Wieless Value As explained in section III. A Simulation Aea 1000 X 1000 m 2 Simulation Time Taffic Model Packet Size No. Of Vehicles 100 Vehicle Speed Packet Rate Radio Popagation Model Routing Potocols 100 Sec 40 connection 512 byte Km/h 4 /Sec Nakagami AODV, AOMDV, DSDV, OLSR

3 B. QoS Metics 1) : The packet dop is counted by the total numbe of when a souce vehicle tansmitted data packet though the netwok to the destination vehicle. The packet dop (P d ) can be counted by Eq. 1. P = P P (1) d s Whee P and P s ae the numbe of and tansmitted espectively. 2) Thoughput: The is counted as a numbe of that have been efficiently sent to the destination vehicles. The (T h ) can be defined as Eq. 2. T = N (2) h t Whee N t is the numbe of data packet bytes in a paticula time. 3) Delay: A paticula packet is tansmitted fom the souce vehicle to the destination and computes the vaiance between sending times and times. The delay (D i ) can be defined as Eq. 3. D = R S (3) i t t Whee R t and S t ae the time of packet and sent. 4) Jitte: The jitte is the vaiance of the packet aival time. The jitte (J i ) can be calculated by Eq. 4. (4) J Di 1 Di = i + C. Othe Pefomance Metics 1) Aveage Thoughput: The amount of data sent by the netwok divided by time peiod. The aveage (A h ) can be calculated by Eq. 5. N 8 A = t (5) h T 100 n Whee T n is the total numbe of vehicles. 2) Nomalized Routing Load (): The total numbe of outing tansmitted pe data packet sent at the destination. The (N l ) is calculated by Eq. 6. N l = R P Whee R p is the numbe of outing in laye 2. 3) Mean Hop: The total numbe of contol o outing fowaded by outing potocol duing the simulation to send data packet deliveed to the destination. The mean hop (M h ) can be defined as Eq. 7. M h = p P f Ps Whee P f and P s ae the numbe outing packet fowaded and sent espectively. 4) Packet Delivey Ratio (): The atio of the data sent to the endpoint to those ceated by the taffic souces. The is calculated by Eq. 8. (6) (7) = P P 5) Routing Cost: It is the atio of outing bytes to taffic packet bytes. The outing (R c ) can be calculated by Eq. 9. R c = N N Whee N and N t ae the numbe of oute & taffic bytes. IV. RESULT ANALYSIS The expeiment is implemented using VanetMobiSim and NS 2.35 an Intel (R) Coe (TM) i7, Windows 7 and Ubuntu platfom to measue the impact of two ealistic mobility on fou VANET outing potocol in an uban scenaio of Dhaka city. A. Quantitative Veification The quantitatively compae the impact of IDM-IM and IDM-LC ealistic mobility using AODV, AOMDV, DSDV, and OLSR outing potocol along with dissimila paamete of QoS metics (dop,, delay, jitte) and othe pefomance evaluation metics (aveage, nomalized outing load, mean-hop, packet delivey atio, outing ) fo infeing the behavio of packet in the dynamic netwok simulation scenaio. The simulation esult is pesented in the following TABLE II-TABLE XVII and Fig. 2- Fig. 9. TABLE II. TABLE III. TABLE IV. TABLE V. t s NUMBER OF PACKET DROP FOR AODV IDM-IM IDM-LC NUMBER OF PACKET DROP FOR AOMDV IDM-IM IDM-LC NUMBER OF PACKET DROP FOR DSDV IDM-IM IDM-LC NUMBER OF PACKET DROP FOR OLSR IDM-IM IDM-LC (8) (9)

4 TABLE VI. THROUGHPUT FOR AODV TABLE XIV., ROUTE COST & MEAN HOP FOR AODV IDM-IM IDM-LC IDM-IM IDM-LC TABLE VII. THROUGHPUT FOR AOMDV IDM-IM IDM-LC TABLE XV. TABLE XVI., ROUTE COST & MEAN HOP FOR AOMDV IDM-IM IDM-LC , ROUTE COST & MEAN HOP FOR DSDV TABLE VIII. THROUGHPUT FOR DSDV IDM-IM IDM-LC TABLE IX. THROUGHPUT FOR OLSR IDM-IM IDM-LC IDM-IM IDM-LC TABLE XVII., ROUTE COST & MEAN HOP FOR OLSR IDM-IM IDM-LC TABLE X., DROP & AVG. THROUGHPUT FOR AODV IDM-IM IDM-LC TABLE XI., DROP & AVG. THROUGHPUT FOR AOMDV IDM-IM IDM-LC Figue 2. Delay fo AODV TABLE XII., DROP & AVG. THROUGHPUT FOR DSDV IDM-IM IDM-LC TABLE XIII., DROP & AVG. THROUGHPUT FOR OLSR IDM-IM IDM-LC Figue 3. Delay fo AOMDV

5 Figue 4. Delay fo DSDV Figue 7. Jitte fo AOMDV Figue 5. Delay fo OLSR Figue 8. Jitte fo DSDV Figue 6. Jitte fo AODV Figue 9. Jitte fo OLSR In TABLE II, TABLE V, TABLE X and TABLE XIV have shown that IDM-IM outpefoms than IDM-LC fo the computed mobility impact of AODV in case of ate (57.47%, 57.94%), (42.53%, 42.06%) and Aveage Thoughput ( kbps, kbps) espectively. In

6 TABLE III, TABLE VII, TABLE XI, and TABLE XV has shown that IDM-LC pefoms bette than IDM-IM fo AOMDV potocol in case of ate (42.30%, 42.78%), (57.70%, 57.22%) and Aveage Thoughput ( kbps, kbps). In TABLE 4, TABLE 8, TABLE XII and TABLE XVI have shown that IDM-LC is pefoming bette than IDM-IM fo DSDV in case of ate (70.09%, 67.42%) and the (29.91%, 32.58%). In TABLE V, TABLE IX, TABLE XIII and TABLE XVII have shown that IDM-IM is pefoming bette than IDM-LC fo OLSR in case of ate (52.92%, 54.23%), (47.08%, 45.77%) and Aveage Thoughput ( kbps, kbps). In TABLE XIV shows that IDM-IM outpefoms among IDM-LC fo AODV with espect to and. In TABLE XV shows IDM-LC outpefoms among IDM-IM fo AOMDV outing potocol with espect to and. In TABLE XVI and TABLE XVII shows IDM-LC outpefoms among IDM-IM fo DSDV and OLSR outing potocols in case of. In Fig. 2 and Fig. 6 has shown that the calculation of delay and jitte fo AODV, IDM-IM pefoms bette than IDM- LC. In Fig. 3, Fig. 4, Fig. 5, Fig. 7, Fig. 8 and Fig. 9 shows that IDM-IM and IDM-LC can pefom well in a cetain time not always fo AOMDV, DSDV and OLSR outing potocols in case of delay and jitte. V. CONCLUSIONS In this wok, two ealistic mobility (IDM-IM and IDM-LC) ae used to show thei impact on fou outing potocols (AODV, AOMDV, DSDV and OLSR) using Nakagami popagation and IEEE p MAC laye with espect to QoS and othe pefomance metics in an uban scenaio. The safety message is peiodically boadcasted using a PBC agent to avoid accident o collision avoidance fo each vehicle. In the simulation esult, it is distinctly suggested that fou outing potocols, and two mobility s wee not up to the mak fo each of the paametes of pefomance metics/qos metics towads the development of ealistic vehicula safety applications. REFERENCES [1] W. Alasmay and W. Zhuang, impact in IEEE p infastuctueless vehicula netwoks, Jounal in Ad Hoc Netwok, vol. 10 (2), pp , Ma [2] J P. Singh, Influences of two-ay-gound and nakagami popagation fo the pefomance of adhoc outing potocol in VANET, Int. Jounal of Compute Applications (IJCA), vol. 45 (22), May [3] S. Azimi, G. Bhatia, R. Rajkuma and P. Mudalige, Reliable intesection potocols using vehicula netwoks, ACM/IEEE Int. Conf. on Cybe-Physical Systems (ICCPS), pp. 1-10, [4] I. Khan and A. Qayyum, Pefomance evaluation of AODV and OLSR in highly fading vehicula ad-hoc netwok envionments, IEEE Mult. Conf. by INMIC'09, Dec [5] V. Godbole, Intelligent dive mobility and taffic patten geneation based optimization of eactive potocols fo vehicula ad-hoc netwoks, Int. Jounal of Infomation and Netwok Secuity (IJINS), Vol. 2 (3), pp , Jun [6] S. Xu, P. Guo, B. Xu and H. Zhou, QoS evaluation of VANET outing potocols, Jounal of Netwoks, Vol. 8 (1), pp , Jan [7] H. Wang, G. Tan and J. Yang, An impoved VANET intelligent fowad decision making outing algoithm, Jounal of Netwoks, Vol. 7 (10), pp , Oct [8] R. Boopathi and R. Piya, Pefomance evaluation of AODV and OLSR in VANET unde ealistic mobility patten, Int. Jounal of Electonics and Communication Engineeing & Technology (IJECET), Vol. 4 (2), pp , [9] M. Jebi, S. Senouci and Y. Ghami-Doudane, Towads efficient geogaphic outing in uban vehicula netwoks, IEEE Tans. On Vehicula Technology, Vol. 58 (9), pp , Nov [10] K. Lan and C. Chou, Realistic Models fo Vehicula Ad hoc Netwok (VANET) Simulations, Int. Conf. On ITS Telecommunications, pp , Oct [11] C. Han, Mehdad Dianati, R. Tafazolli, R. Kenchen, and X. Shen, Analytical study of the IEEE p MAC sub-laye in vehicula netwoks, IEEE Tans. On Intelligent Tanspotation Systems, Vol. 13 (2), , Jun [12] V. Cabea, F. Ros and P. Ruiz, Simulation-based study of common issues in VANET outing potocols, 69 th Vehicula Technology Confeence, [13] S. Gwalani, E. Belding-Roye and C. Pekins, AODV-PA: AODV with path accumulation, Int. Conf. On Communications, Vol. 1 pp , May [14] X. Du, Y. Wang, J. Ge and Y. Wang, A method fo secuity enhancements in AODV potocol, 17 th Int. Conf. Poc. On Advanced Infomation Netwoking and Applications, pp , Ma [15] A. Chintawa, M. Chattejee and A. Vidhate, Pefomance analysis of ad-hoc on-demand multipath distance vecto outing potocol with accessibility and link beakage pediction, 2 nd Int. Conf. And Wokshop on Emeging Tends in Technology (ICWET), [16] S. Mallapu and S. Tedal, Enhanced ad-hoc on-demand multipath distance vecto outing potocol (EAOMDV), Int. Jounal of Compute Science and Infomation Secuity (IJCSIS), Vol. 7 (3), Ma [17] F. Diamantopoulos and A. Economides, A pefomance study of DSDV-based clustepow and DSDV outing algoithms fo senso netwok applications, Int. Symposium on Wieless Pevasive Computing, Jan [18] T. Clausen and P. Jacquet, Optimized link state outing potocol fo adhoc netwoks, Oct [19] M. Fioe, J. Häi, F. Fethi, and C. Bonnet, Vehicula mobility simulation fo VANETs, Annual Simulation Symposium (ASS), Ma