Australian Journal of Basic and Applied Sciences. Advanced Heterogeneous Vehicular Network for Emerging Vehicular Services

Transcription

1 AENSI Journal Autralian Journal of Baic and Applied Science Journal home page: Advanced Heterogeneou Vehicular Network for Emerging Vehicular Service 1 Saied M. Abd El-atty and KontantinoLizo 1 Department of Computer Science and Information, Art and Science College, Salman Bin Abdulaziz Univerity, Wadi Adwair, Riyadh, Saudi Arabia. 1 The Dept. of Electronic and Electrical Communication Engineering, Faculty of Electronic Engineering, Menoufia Univerity, 32952, Menouf, Egypt. 2 Department of Informatic, Faculty of Mathematic and Natural Science, Univerity of Olo, Norway. ARTICLE INFO Article hitory: Received 24 November 2013 Received in revied form 18 January 2014 Accepted 29 January 2014 Available online 25 February 2014 Key word: Failure Connectivity Probability; Failure Acce Probability; MAC; V2V; V2R; AHVN ABSTRACT Advanced heterogeneou vehicular network (AHVN) i a promiing architecture for providing vehicular ervice in the next generation of vehicular network. AHVN i an integrated architecture between vehicular ad hoc network and exiting cellular wirele network. In thi work, we propoe a Multihop vehicular connectivity model in V2V ytem, which depend on the phyical characteritic of the roadway and fale hop initiation connectivity. Then, we determine the failure probability of vehicular connectivity in V2V ytem. Baed on interoperability utility, we employ the failure connectivity probability a a handover criterion to communicate with V2R network. Subequently, we propoe an efficient medium acce control (MAC) method baed on collaborative code for reource management in AHVN. A a reult, we determine the failure acce probability by employing a Markov chain model. The analyi of the propoed MAC in term of tranmiion capacity, delay and acce failure probability i driven. The numerical and imulation reult demontrate the effectivene of the propoed framework AENSI Publiher All right reerved. To Cite Thi Article: Saied M. Abd El-atty and KontantinoLizo., Advanced Heterogeneou Vehicular Network for Emerging Vehicular Service. Aut. J. Baic & Appl. Sci., 8(1): , 2014 INTRODUCTION Vehicle to vehicle (V2V) and vehicle to roadide infratructure (V2R) communication are uually integrated to deign a ubiquitou, cooperative communication ytem, baed on the coverage area. V2V communication involve vehicular node on the road that form a vehicular ad hoc network (VANET). However, V2R involve vehicular node, roadide bae tation and acce point connected with Internet. On the other hand, intelligent tranportation ytem (ITS) are poied to play a fundamental rule in introducing the updated and mot uitable olution for afety application, while non-afety ervice engage communication in with the commercial wirele network (GSM, UMTS/HSPA and WiMAX/LTE). In order to upport variou afety and non-afety ervice with the deired quality of ervice (QoS), vehicle in vehicular network hould not only follow pecific protocol, but alo cooperate with each other, to enhance the overall network performance and efficiently utilize reource (Chen and Cai, 2005) and (Lequerica et al.,2010). Therefore, the idea of combining V2V baed ytem with the current cellular ytem i a feaible olution not only for improving the utainability of the ITS (particularly for afety related application), but alo for increaing the throughput. The vehicular communication have recently enviioned interet from both academia and indutry. The main important cooperation related parameter between V2R and V2V i calculating the parameter of the communication path known a dedicated hort range communication (DSRC) cot. However, DSRC technology will only be effective when it i ubiquitouly deployed but thi will not happen due to the retriction of DSRC deployment in paenger vehicle. On the other hand, IEEE p tandard have iued the wirele acce in vehicular environment (WAVE) to upport vehicular ervice (Rita, 2010) and (Alamary and Zhuang., 2012). Mot tudie in the area of vehicular network only provide a ingle type of ervice in uch network and are only baed on a non-cooperative radio acce network. Therefore, the integration of exiting wirele acce technologie with ITS enable a plethora of ervice, uch a accident avoidance, emergency call, traffic congetion control, vehicle entertainment, Internet acce and mobile buine. Correponding Author: Saied M. Abd El-atty, Art and Science College, Salman Bin Abdulaziz Univerity, WadiAdwair, Riyadh, Saudi Arabia. Phone: the; .oliman@au.edu.a or abdelatty@gmail.com

2 309 Saied M. Abd El-atty and KontantinoLizo, 2014 The advanced heterogeneou vehicular network (AHVN) i a promiing candidate network for next generation vehicular network (Ekram et al., 2010). AHVN deploy multiple radio and multiple acce technologie in a cooperative communication manner and thereby vehicular ervice can be upported through V2V and V2R integration. V2R communication involve vehicular node and road ide bae tation. In equel, each vehicle can end or receive data to/from bae tation (BS). However, if a vehicle could not directly end it data to BS, it can relay it data to other vehicle until the data reach the BS uing Multihop path tranmiion. In addition, by deploying AHVN, the vehicle are able to requet ocial ervice with different requirement in term of delay, bandwidth, error rate, coverage area, etc. at any time and any place. Current acce technologie uch a wirele LAN (WLAN IEEE a/b/g/n/p tandard), WiMAX (IEEE a/e tandard), ultra wideband (UWB IEEE a tandard), 3G and 4G cellular mobile network are deigned for pecific ervice requirement (Ekram et al., 2010l). Without lo of generality, we deploy a 4G wirele technology (backward compatibility with 2G, 3G) for vehicular ervice that provide a very broad coverage, upport high-mobility vehicle and moother handover. However, the capacity of BS may be diminihed when the code colliion happen. Subequently, the management of the wirele acce in AHVN i alo a challenge iue. In thi tudy, we firtly propoe a Multihop connectivity model in order to determine the failure probability of vehicular connectivity (P f ) in V2V ytem, P f depend on the probability of fale hop connectivity, headway ditance and received ignal trength (RSS). Secondly, baed on the interoperability utility, we employ the failure connectivity probability a a handover criterion to communicate with V2R. Next, we propoe an efficient medium acce control (MAC) collaborative-baed code cheme for reource management in AHVN. Finally, we developed a Markov chain model in order to determine the probability of failure acce to V2R ytem. The remainder of the paper i organized a follow. The related work i dicued in Section II. A vehicular communication framework in AHVN which include the interoperability mechanim between V2V and V2R ytem i preented in Section III. In Section IV, we preent the numerical and imulation reult that evaluate the performance of the propoed framework. Finally, a brief concluion of thi work i given in Section V. The mathematical notation and their phyical meaning that are ued in thi paper are illutrated in Table I. Table I: Mathematical notation and phyical meaning Mathematical Phyical meaning notation μ Probability of market penetration rate L,W Roadway i egmented into microcell with length (L) and lane width (W) V i, V o, V n Current vehicle, old vehicle, new vehicle in V2V multi-hop ytem f v(v ) Probability denity function (pdf) of vehicle velocity, v l v, d Ditance headway and a portion of ditance headway at point P, repectively Q v,d v Volume of vehicular traffic, vehicular denity f θ(θ) The pdf of V i direction of motion θ τ Broadcat ignaling delay time M v(t), K(t) Number of vehicle in [0, L] at time t,the total number of vehicle that arrive at the time interval [0, t] λ v, m v Vehicle arrival rate and number of vehicle R 0 Maximum tranmiion range of each vehicle P c Probability of vehicular connectivity P fale Fale probability of initiation hop P f Failure probability of vehicular connectivity φ,φ th,φ min Received ignal trength (RSS), threhold of φ, minimum value of φ T Slot duration of packet tranmiion υ Probability that a vehicle end one packet in a ingle lot p j Probability of j licened vehicle are active at a given time lot C Available code in BS =1/C Probability that an active vehicle elect a certain code. P Probability of ucceful tranmiion for an active vehicle I(P) Denote to the interoperability function for the event of failure probability (P f) of vehicular connectivity in V2V and g Tranmiion capacity and the gain in V2R network repectively. P AF Acce failure probability in V2R network. D Average delay for ucceful packet tranmiion Relatedwork: Intelligent tranportation ytem (ITS) have iued the tandard for vehicular communication, according to inter-vehicle communication (IVC) technology. Thi type of communication i dynamic by nature due to the vehicle mobility and network topology, i.e. called vehicular ad hoc network (VANET) or vehicle to vehicle V2V, enabling the deployment of Vehicular Social Network. The connectivity probability in V2V ha been tudied through many reearch work by analytical and imulation model (Babu and MuhammedAjeer., 2011) and (ANSI-IEEE Standard., 1999). However, emerging wirele ervice uch a ocial ervice in vehicular network make the medium acce of V2V ytem not uitable to achieve the QoS requirement of

3 310 Saied M. Abd El-atty and KontantinoLizo, 2014 uch ervice. Therefore, the integration architecture of V2V and the exiting wirele acce technology would contribute to better acce ervice and to achieve the QoS requirement. In (Ekram et al., 2010) the author preented different architecture of advanced heterogeneou vehicular network (AHVN) for vehicular telematic ervice. They lightened the major challenge and approache for deigning the AHVN uch a wirele acce trategie, network election, MAC protocol, etc. The author in (Taleb and Benlimmane., 2010), propoed the baic guideline deign for network integration architecture of V2V and 3G network. The main purpoe of thi pivotal reearch i to elect only a minimum number of vehicle to communicate with 3G network a gateway. In equel, the author in (Benliman et al., 2011) introduced integration architecture between V2V and 3G network by deploying clutering gateway. They made ue of UMTS received ignal trength (RSS) metric of the vehicle intead of vehicle velocity variation for dynamic clutering mechanim. On the other hand, the author in (Chun et al., 2011) developed an analytical model to fully characterize the acce probability and connectivity probability performance for infratructure-baed vehicular relay network. Thi reearch i capable of improving acce and connectivity probabilitie in V2R and V2V baed relay network. Furthermore, in term of MAC protocol for V2V ytem, there are many work that have been invetigated (CarTalk, 2000) and (FleetNet, 2001). Recently, the author in (Luan et al, 2010) propoed a MAC mechanim baed on ditributed coordination function (DCF) for V2R communication. They developed an analytical model to evaluate the performance of DCF in high-peed vehicular environment. Conequently, the author in (Sepulcre et al., 2011) propoed a contextual cooperative congetion control policy between vehicle and infratructure node that exploit the traffic information of each vehicle to reduce the unneceary interference. In addition, the author in (Zhou et al., 2011) propoed a novel adaptive ditributed cooperative relaying MAC protocol for vehicular network, thi reearch addreed the inherent problem of MAC baed IEEE tandard. A a popular and emerging wirele tandard, the Vehicular Social Network (VSN) preent a qualified framework, for the formation of trut-oriented communitie among moving vehicle. Since thi technology i fairly new, the related work i carce and inadequate but with great interet and impact for urban environment. In (Abbani et al., 2011), a model i preented to cope with the adminitrative-related iue through a dynamic trut management module. Thi model i analyzed mathematically to ae the probability of deceitful penetration (Node-to-Node, Node-to-Group) while exhibiting a low-latency ratio Latency i meaured by the level of interaction and the repone that occurred. Reource allocation in VSN require enitive balancing, ince it mitigate the complexity of wirele network in decentralized vehicle tructure with potentially heterogeneou diverified traffic requirement. More pecifically, in (Lequerica et al, 2010) the challenge of VSN in thi context are ummarized: Effortle human intervention (by utilizing Human-Machine Interface (HMI) and command execution through peech channel) Adaptation/penetration of upported capabilitie over diverified hardware and application heterogeneou device Anonymity and privacy Interoperability in reference to collecting and adminitrating information Efficient ue of network reource Authentication / Security related iue Privacy and ecurity in particularare extenively covered in (Algana et al., 2011) for ocial-aware vehicle communication. Although the cenario i both feaible and ucceful, the preence of truted uper-entity, centralizing audit mechanim, i eential. In (Fei et al, 2011), a cheduling policy i introduced and analyzed to effectively integrate the unremitting iue of ocial need into lower network module. The author examine the Centrality Degree i.e. an entity i more central than other if it i more viible /prominent to other entitie. Moreover, the propoed cheduling olution encompae two additional factor: the level of ervice requet from each node and the eparate ervice type that requet i compried of to derive a cheduling deciion. To the bet of our knowledge, no reearch dicued the failure connectivity probability (P f ) in V2V ytem that i caued by different parameter uch a the phyical characteritic of the roadway, vehicular traffic denity, broadcat ignaling delay, fale hop initiation connectivity. Moreover, the interoperability mechanim for employing the failure connectivity probability a handover criterion for etablihing a communication link to V2R i propoed. Framework For Vehicular Connectivity In Ahvn: The framework for vehicular connectivity in AHVN i mainly baed on the interoperability between V2V and V2R network. A implified architecture for the AHVN i illutrated in Fig.1.

4 W L 311 Saied M. Abd El-atty and KontantinoLizo, 2014 Internet Core Network BS2 BS1 Multihop path 3 rd hop Vd detination Direct hop path 2 nd hop V1 BS0 1 t hop ource V Fig. 1: A implified architecture of AHVN. The baic component of the network infratructure are V2V ytem baed-multihop trategy, BS, core mobile network and Internet. The deployment of AHVN enable vehicle to get acce to BS outide their tranmiion range via tranmiion trategie, ingle hop or Multihop tranmiion, which improve the network coverage (Ekram et al., 2010). Further, cooperative ue of the multiple radio enhance the performance of the network by allowing more parallel tranmiion in different frequencie. The connectivity over V2V network depend on inter-vehicle connectivity (IVC) bai and Multihop tranmiion trategy. In addition, the vehicle can directly connect through a ingle hop path a long a it location inide the coverage area of the erving BS. On the other hand, the vehicle can ue a V2V Multihop path connectivity to communicate with BS during it movement from one cell egment to other (Sou and Tonguz., 2011). A. Interoperability utility between V2V and V2R ytem: In order to provide interoperability communication between V2V with V2R ytem, two important concern need to be effectively addreed.thi functionality could be accomplihed with the ability of vehicle to elect the appropriate acce network during it roaming from one cell to the other. Firtly, the vehicle may etablih a Multihop connectivity link by uing DSRC technology in V2V network according to phyical characteritic of roadway uch a headway ditance, vehicular traffic denity, hop fale initiation connectivity and received ignal trength (RSS). However, thi connectivity may fail due to unavailability of phyical link in V2V network; hence the probability of failure connectivity (P f ) in V2V network hould be etimated. The next claim, the received ignal trength (RSS) at vehicle will be deteriorated; thereby we make ue of the interoperability utility and the probability of failure connectivity (P f ) in order to acce collaborative code in roadide infratructure (V2R) to get the vehicular ervice or Internet ervice.therefore, we compute the ucceful tranmiion probability in V2R network taking into account the probability of failure connectivity (P f ) in V2V network.the algorithm of interoperability communication between V2V and V2R ytem can be concluded a follow: Algorithm 1 - Failure Connectivity in V2V Network 01 For each vehicle V i that ak a connectivity link in a particular cell of roadway with length L& width W, do 02 Meaure the market penetration rate (μ), vehicular traffic denity D v, ditance headway (l v) 03 According to V i moving direction (θ) 04 Etimate vehicular velocity v 05 Uing DSRC technology, maximum tranmiion range (R 0) 06 Determine the probability of connectivity (P C), otherwie 07 If no connectivity availability in V2V network 08 Meaure fale hop initiation probability 09 Then uing the previou meaurement to find optimal headway ditance

5 312 Saied M. Abd El-atty and KontantinoLizo, Meaure the received ignal trength (RSS), φ 11 If (φ>φ th) 12 Succeful connectivity 13 Ele 14 Failure connectivity 15 Compute the failure probability of vehicular connectivity (P f) which will be ued in V2R ytem Algorithm 2 Acce Failure in V2R Network 01 V i fail to have a connectivity link in V2V network, with P f 02 I(P) i the interoperability function baed on RSS 03 In V2R network, licened vehicle, m v can acce collaborative code, C 04 Apply random method for code election 05 Colliion may occur if another vehicle end data by uing the ame code. 06 Deploy Markov chain to repreent the wirele channel tate 07 Determine the probability of ucceful tranmiion, i.e. the probability of licened vehicle electing a unique code from the available code C. 08 Compute the acce failure probability B. V2V ytem: 1. Mobility model: In inter-vehicle connectivity (IVC) ytem, vehicle equipped with wirele communication device along roadway etablih V2V ytem (Zeadally et al, 2012). The feaible connectivity depend on the vehicular traffic denity on the roadway. The probability of whether or not a vehicle equipped with wirele communication device i random with the probability of market penetration rate, denoted by μ or (1- μ) repectively. We conider the roadway, egmented into microcell with n up and n down lane, and we aume each lane with length L and width W, a hown in Fig.2. The movement of the vehicle in each lane i conidered to be in the ame direction. We conider the cenario in which a vehicle (V i ) i currently connected by old vehicle (V o ). V i i moving with a pace mean velocity v. Let v be uniformly ditributed in [v min, v max ], therefore the probability denity function (pdf) of v can be expreed a follow: 1 fv( v) =, vmax < v < v (1) min v v max min There i alway a relationhip between the mobility of vehicle and the network connectivity of the ad hoc ytem. For example, during it coure of movement in a typical lane, vehicle V i i connected with old vehicle V o and hop a new communication path with new vehicle V n new hop path old hop path V o V i V n θ 1 l v d P W L Fig. 2: A egment of highway microcell tructure. Ditance headway (l v ) i a key parameter in calculating the overall capacity of vehicular ad hoc network [17]. Therefore, ditance headway i related to vehicular traffic denity (D v ) a follow 1/D v. The relationhip between the volume of vehicular traffic Q v and denity i given by Q v =D v v,where v i a function of the vehicular velocity through the following relation, v = v f [1-D v / D jam ],where v f i the mean velocity for free flow condition and D jam i the denity under jam traffic. Under free flowing, all vehicle are able to move at the velocity characteritic of the roadway, a determined by it phyical feature, velocity limit, traffic light, etc. When vehicle V i reache the point P at ditance (d) of the ditance headway (l v ) then it probably hop with the new V n, during the deirable ditance headway before being out of tranmiion range. When V i i located at point P, we aume that it can move in any direction with equal probability, therefore the pdf of V i direction of motion θ can be expreed a follow: 1 f θ ( θ) =, π < θ < π (2) 2π Without lo of generality, we conider the direction and velocity of V i remaining the ame from point P till it move out of the coverage area of new vehicle V n. A the ditance headway to new vehicle i mall, thi aumption i realitic. For example, for a vehicle moving at 100 km/hr conidering the broadcat ignaling

6 313 Saied M. Abd El-atty and KontantinoLizo, 2014 delay time (τ ) i 2 ec, d will be 50m. Hence during high vehicular traffic denity, the velocity and alo the broadcat ignaling delay will be decreaed; therefore d will be much maller. Vehicular connectivity: In a free-flow traffic highway, we are intereted in thoe vehicle that fall in the interval [0, L] a illutrated in Fig.II, the number of vehicle at time t i a random variable. Let M v (t) denote to the number of vehicle in [0, L] at time t, and K(t) denote the total number of vehicle that arrive at the time interval [0, t]. Let the vehicle arrival rate be λ v, then we have: k ( λv ) v t P{ K() t k} t λ = = e (3) k! The ditance headway between two conecutive vehicle i exponentially ditributed, i.e., l v i exponentially ditributed with the following cumulative ditribution function a follow (Schonhof et al., 2006): F ( l ) 1 e 1 e cdf v lv. Dv mv.( lv / L) = = (4) The ditribution probability of M v (t) can be derived a follow: mv ( Lλv / v ) Lλv / v P{ Mv() t = mv} = e (5) mv! Thereby, M v (t) i Poion ditributed with mean λ=lλ v /v. Thi i very important for the ytem deigner in order to expect the connectivity probability according to vehicle ditribution a well a tranmiion range. Accordingly, auming an ad hoc network with m v vehicle Poion ditributed over a egment with length L and maximum tranmiion range of each vehicle R 0. The hop path between two vehicle can be etablihed if ditance headway i equal to or le than R 0. Then the probability that all m v vehicle are connected can be expreed a follow (Schonho et al., 2006): P = Pr{ l R, l R,..., l R } (6) c m v 1 0 In reality the ditance headway l i i an independent and identically ditributed (iid) random variable, therefore from (4) and (6), P c i given by c mv 1 Pr{ i 0} (1 Dv. R0 mv 1 ) (7) i = 1 P = l R = e The failure probability of vehicular connectivity: A ucceful communication between vehicle i etablihed when a broadcat ignaling for initiation connectivity i received before a complete connectivity i executed. A hown in Fig.2, it i illutrated that the initiation of hop communication to the new vehicle (V n ) arie only if V i ' direction of motion from P within the range {-θ 1, θ 1 }, where θ 1 =arctan(w/2d). Otherwie, the fale hop initiation probability i a fale one. Therefore, by uing (2), the fale hop initiation probability can be expreed a follow: θ1 1 W Pfale = f θ ( θ) dθ = 1 arctan (8) π 1 2 d θ Themoving direction of V i from P i β {-θ 1, θ 1 }, the time t ittake to move out of the coverage area of the old vehicle (V o ), therefore, the pdf of t can be derived a follow: d d d, < t < 2 2 f t () t = 1 ( ) v (9) θt vt θ1t ( vt ) 0 otherwie The failure probability of vehicular connectivity, P f i expreed a: W 2 2 ( 2 ) + d 1, τ > v (10) W 2 2 d ( 2 ) + d Pf = pt ( < τ), < τ < v v d 0 v τ

7 314 Saied M. Abd El-atty and KontantinoLizo, 2014 d v whereτ i the time delay of ignalling broadcat for connectivity and p(t<τ) i the probability that t<τ when < τ < ( ) W 2 v d,uing (9) in order to obtain p(t<τ) a follow: τ d 1 d p( t < τ ) = dt arcco 2 2 d πt ( vt) d θ v τ v (11) P f Now uing (10) and (11),P f i given by W 2 2 ( 2 ) + d 1, τ > v W 1 d d ( 2 ) + d = arcco, < τ < θ vτ v v d 0 τ v 2 2 (12) Outage probability: The meage hop for one vehicle to next vehicle within the ame driving direction (Schonhof et al., 2006). According to the number of the vehicle equipped with wirele device, that i the market penetration rate (μ), therefore if there i a receiver vehicle in the broadcat range of ending vehicle, a meage hopping i poible. For ucceful meage hopping, the ditance headway mut be maller than the maximum tranmiion range; otherwie, the meage hopping fail. The probability of finding a receiver vehicle for meage hopping i baed on the maximum tranmiion range (R 0 ) i given by 0 R0 lv ( µ. Dv ) v R0 ( µ. Dv ) (13) 0 p( l < R ) = e dl = 1 e v Therefore, the connectivity probability over a ditance (for example, r =1km) in one directional roadway i baed on the ucceful probability of meage hopping in IVC ytem and it i given by (CarTalk, 2000): r R0 i 1 ( µ. Dv ) i 1 0 i= 1 ( i 1)! (14) P () r = 1 ( r ir ) c ( µ. Dv ) ir0 (. Dv ) 1 ( r ir 0) e µ + i where r/r 0 i the larget integer maller than or equal to r/r 0. Therefore, the outage connectivity probability (P outage ) can be defined a the event that there i at leat one vehicle diconnected from the remaining of the Multihop vehicle in the network. Hence P outage can be expreed a follow: P () r = 1 P() r (15) outage c Optimal headway ditance: The hop initiation for connectivity with the new vehicle V n i baed on the received ignal trength (φ). In equel φ depend on the optimal ditance for ucceful connectivity. Let the threhold of received ignal trengthi φ th and the minimum value of received ignal trength i φ min, required for ucceful connectivity between V i and V n. The optimal value of d can be etimated for a deired value of P f. Equation (12) i a nonlinear equation of d. A cloed form expreion may not be alway poible. However, an approximate value of d can be etimated uing: 1 d π d co v 2 v P τ τ (16) f = 1 W π 2d tan d (2 d) + W Numerical method can be ued to determine d. We employ the Biection numerical method to olve for d. Hence, calculation of d doe not have much computational complexity and can be eaily implemented in the vehicle' radio oftware. Fig.3 how the numerical reult for olution Eq.(16), the relationhip between the optimal headway ditance and the failure probability of connectivity at different value of τwhen velocity

8 315 Saied M. Abd El-atty and KontantinoLizo, 2014 iv =80km/hr. For example, a illutrated in the Fig.3 at τ=3ec and a threhold value of P f =0.4, the optimal headway ditance for ucceful connectivity in V2V i approximately le than or equal to 60m. Once the optimal ditance (d 0 ) i determined, the correponding value of RSS i calculated uing the path lo model a follow [19]: α d 0 = Pr d0 + Pr() l ( ) σ (17) l wherel i the ditance between V i &V n and P r (d 0 ) i the received power at a known reference ditance d 0, it can be obtained for line of ight (LOS) or non-line of ight (NLOS) vehicular environment. αi the path lo exponent, it typical value range from 2 to 8 in a vehicular environment (Wang et al., 2009). Finally σ i zero mean Gauian random variable. The typical value of σ i 8dB.The value of σ depend on the lane ize (W). Therefore, received ignal trength (φ) value when V i at a d ditance from V n i given by φ = 10log [ PW ( d)] (18) 10 r Fig. 3: Optimal headway ditance veru P f at different value of τ. C. Vehicular to roadide infratructure (V2R) model: 1. Collaborative code: In each BS, we aume that there are C collaborative code for excluively ue for m v licened vehicle (where C < m v ). We conider m v vehicle uing the ame coding technique but at different time lot. When the failure probability (P f ) of vehicular connectivity in V2V increaed above a threhold value, a licened vehicle i able to communicate with the BS for code acce by uing Multihop tranmiion trategy. A licened vehicle may randomly elect one of the available collaborative code (C) and end it data traffic. A maximum of C vehicle can utilize the channel imultaneouly, if they have ued different code, otherwie colliion may occur. 2. MAC baed collaborative code: In vehicular to roadide infratructure (V2R) communication, MAC protocol define how the vehicle and the bae tation (BS) hare common radio code for vehicular communication. On the other hand, in V2V communication, MAC protocol define how the vehicular node hare code to tranmit data among them (Ekram et al, 2010). However, due to the vehicle mobility frequent handover and the high variant of topology along the road; an efficient MAC protocol i a challenging iue. Therefore, the integrated architecture of AHVN enable vehicle to get acce to the cellular ytem (BS) outide their tranmiion range via Multihop tranmiion trategy. In thi ection, we propoe a medium acce cheme baed-collaborative code for the reource management in AHVN. Thi cheme favor the vehicle that failed to have a communication link in V2V; they will have the ability to acce the V2R network. We conider all licened vehicle to have equal priority and they are able to tranmit packet with equal length at any time they want. We aume that the lot duration of packet tranmiion i denoted by a contant value of T. The probability that a vehicle end a packet in one lot iυ. In any cell egment, the probability of j licened vehicle, being active at a given time lot can be expreed a follow:

9 316 Saied M. Abd El-atty and KontantinoLizo, 2014 p j m j = υ [1 υ] v j v m j (19) When licened vehicle are active, the probability of ucceful tranmiion (j) i the probability of each vehicle electing a unique code from the available code C, i given by 1, j = 1 C! j (20) ( j) = γ, 1< j C ( C j) 0 j C where =1/C i the probability that an active vehicle elect a certain code. The probability of ucceful tranmiion, P for any number of active vehicle i then given by: C P = ( j) p (21) j j = 0 Baed on the above aumption, the wirele channel can be in one of three tate: VANET fail, collided or ending. In our analyi, we utilize Markov chain to repreent the wirele channel tate of vehicle a hown in Fig.IV. State 1 repreent the wirele channel tatu with failure connectivity probability (P f ) in V2V ytem, with probability I(P). State 2 decribe the colliion tatu with probability [1-P -I(P)]. Finally, tate 3 repreent the ending mode with probability P. According to the Markov chain illutrated in Fig.4, we can obtain the tate tranition matrix a I( P) I( P) I( P) 1 I( P) P 1 I( P) P 1 I( P) P (22) Ρ= P P P wherei(p) denote to the interoperability utility for the event of failure probability (P f ) of vehicular connectivity in V2V ytem, i.e. I(P) i a function of received ignal trength ( ), it i given by: Pf φ < φth I( P) = (23) p0 φ φth wherep 0 i the tate probability of idle channel and the equilibrium ditribution vector ( ) can be derived a follow: I( P) = 1 I( P) P (24) P The tranmiion capacity (η) of the V2R network i given by C η = ( ) j ( j) p (25) j j = 1 Additionally, the gain (g) and acce failure probability (P AF ) in V2R network repectively can be meaured by taking a SALOHA a reference bai, in the following manner: η g = 20log 10( ), ηsaloha (26) η P AF = 1 C υ (27) where η i the probability of a ucceful tranmiion for a given licened vehicle. On the other hand, C υ the efficiency of the MAC (ε) can be determined a follow: η η ε = = (28) average input traffic mvυ Alo, we conider the average duration of the random wait period to be τ p, hence the average delay for ucceful packet tranmiion can be expreed a follow: p mv D τ = υ (29) η

10 317 Saied M. Abd El-atty and KontantinoLizo, 2014 I(P) 1 VANET fail 1-I(P)-P P I(P) I(P) P 3 2 P Sending Fig. 4: Markov chain of vehicular wirele code tate 1-I(P)-P 1-I(P)-P Collided Numerical And Simulation Reult: The performance evaluation of the propoed framework i evaluated for a highway a a cae tudy. Thereby, we firtly tudy the performance of the V2V ytem, then, according to the interoperability algorithm, we can obtain the behavior of P f to tudy the performance of the V2R ytem. A. Performance Evaluation in V2V ytem: In thi ection, we tudy the performance of the V2V ytem model in term of fale initiation probability, connectivity probability and failure connectivity probability. The characteritic of vehicular environment uch a D v, μ, v and R 0 are et to default parameter. A vehicular traffic denity in a cell egment (L=1km) of highway i given by (D v =m v /L).The numerical reult of the V2V ytem model are hown in the following figure. Fig. 5: Multi-hop connectivity probability veru vehicular traffic denity Fig.5 illutrate the probability of Multihop connectivity for increaing D v when μ et to 10 per cent at different value of R 0. A expected, the connectivity probabilitie at high tranmiion range are higher than the connectivity at low tranmiion range. The importance of thi curve lie in the fact that the deigner of Multihop connectivity in vehicular network ha the capability to preciely pre-ae the number of vehicle per highway egment length for a deired level of connectivity. Due to the phyical characteritic of the highway, the aociation between probability of fale hop initiation connectivity and the portion of ditance headway (d) for different lane ize, W i hown in Fig.6. A we can ee, P fale increae when the ditance headway increae at different value of lane ize. In addition for each particular value of W, P fale i ignificantly reduced at large value of W. Thi tem from the fact that the number of vehicle, occupying thoe lane ha probably elevated. Notably, it how that the problem of fale initiation connectivity become more and more evere when lane ize decreae.

11 318 Saied M. Abd El-atty and KontantinoLizo, 2014 Fig.7 how a correlation between the failure connectivity probability (P f ) and mean pace vehicular velocity (v ) for different value of d when τ =3ec. We can ee that a v increae, P f ignificantly increae. Thi figure alo how that for a large value of d (i.e., ditance headway i mall), P f decreaed. In other word, when the vehicular traffic denity i high, then the ditance headway i mall; in equel P f i ignificantly reduced. Obviouly, the higher the number of equipped vehicle V2V on the highway ector i, the higher the occurrence of having a Multihop connected V2V network. Fig. 6: Fale initiation probability veru ditance headway The ucceful meage hopping probability between the equipped vehicle through IVC i not only baed on the maximum tranmiion range but alo onthe vehicular traffic denity. The outage connectivity probability veru number of vehicle in a ector L=1km with μ et to 25 per cent, at different value of R 0 i illutrated at Fig.8. A we can ee in thi figure, the outage connectivity probability i ignificantly diminihed by increaing the number of vehicle in the highway ector. Furthermore, the outage connectivity probability become more and more reduced at high value of maximum tranmiion range. B. Performance Evaluation in V2V ytem: At BS, we aume that the number of collaborative code (C) i 2, 3, and 5 code for excluive ue by licened vehicle. In addition, baed on the interoperability algorithm, the handover criterion to V2R i introduced. In the following numerical reult, we adopt the failure probability of connectivity in V2V ytem et to 0.4 a threhold value for handover criterion to V2R ytem. Fig. 7: Failure connectivity probability veru vehicular velocity Further, we develop a imulation model according to the main parameter of n-2 imulator for vehicular network (Murray et al.,2010). The imulation cenario i ued to keep the tractability of the analytical model.we conidered a continuou highway of length 10 km with 10 cell of length 1km. Vehicle are

12 319 Saied M. Abd El-atty and KontantinoLizo, 2014 generated following a Poion proce with arrival rate 0.3 and 0.5 veh/ec and the vehicle peed i uniformly ditributed with {20, 120}km/hr. We aume that the number of vehicle per cell (m v ) i 50. Fig. 8: Outage connectivity probability veru vehicle number in ector L=1km In order to implify the tudying of the performance MAC cheme, we apply a real-time ocial traffic flow uch a voice chat with the following aumption: 1. The ytem upport m v =50 vehicle. 2. Voice data rate i 64kbp during activity period. 3. The payload of voice packet i 160 bit/packet; hence voice packet rate i 400packet/ec. 4. The total packet ize i 200 bit including the header, FEC encoding, and payload. 5. The wirele code tranmiion rate i 10Mbp, thu packet tranmiion delay i 0.02m. 6. Maximum threhold packet delay i 3 mec. Therefore, from the above aumption, we conider a ytem with time lot T=0.02m, τ p i equal to one time lot and the average input traffic i expreed by m v υ. We now preent a naphot of our imulation model upported by numerical reult in order to evaluate the propoed framework in AHVN. Fig. 9: Tranmiion capacity v. the average input traffic Fig.9 how the tranmiion capacity (η) performance a the average input traffic increae. We can ee in thi plot, we can oberve that ee the curve of the tranmiion capacity i getting better when the number of collaborative code increae; becaue when the number of the collaborative code increaed, the vehicle'

13 320 Saied M. Abd El-atty and KontantinoLizo, 2014 ucceful tranmiion probability increae. Furthermore, it i clear that our analyi baed Markov chain can evaluate preciely the performance of the propoed MAC. The behavior of the curve indicate that the tranmiion capacity i teadily increaed from low to medium traffic load and the tranmiion capacity goe down at high traffic load. Fig.10 how alo the performance of average packet delay at different value of C by increaing the average input traffic. A we can ee, the average packet delay i ignificantly reduced when the number of the collaborative code i high. Subequently, the higher the number of C, the higher the gain of ytem i, a hown in Fig.11. Fig. 10: Average packet delay v. the average input traffic C. Performance comparion: In thi ection, we introduce the validation of the propoed cheme. We compare the performance of the propoed cheme with adaptive ditributed cooperative medium acce control (ADC-MAC) (Zhou,2011) and the IEEE protocol (ANSI-IEEE Standard., 1999). ADM-MAC i accomplihed through adaptively electing the mot uitable relay or helper and the tranmiion mode in accordance with the code quality and location of the relay vehicle. In addition, ADM-MAC i alo compatible with the tandard IEEE protocol that depend on the Ditributed Coordination Function (DCF) protocol. We adopt our ytem model to perform the ytem analyi of the ADM-MAC and IEEE protocol. We oberve that by increaing the input average traffic in our ytem model reult in packet ize expanion in ADM-MAC cheme and IEEE protocol. Fig.12 how the tranmiion capacity performance comparion between our cheme and ADM-MAC / IEEE cheme. A hown in thi figure, it i traightforward that the behavior of the propoed cheme i comparable with both ADM-MAC and IEEE protocol. In addition, the tranmiion capacity of the propoed cheme outperform both ADM-MAC and IEEE cheme. Fig. 11: Gain veru the average input traffic

14 321 Saied M. Abd El-atty and KontantinoLizo, 2014 Fig. 12: Tranmiion capacity v. the average input traffic To ummarize, our framework analyi how that, the vehicular connectivity can be improved by imple integration architecture between V2V and V2R network a a implified form of AHVN. A hown in the outcome, the AHVN network tranmiion capacity i increaed by increaing the number of collaborative code in V2R. Subequently, acce failure and average delay are ignificantly decreaed. Concluion And Future Work: We preented a framework for emerging vehicular ervice in advanced heterogeneou vehicular network (AHVN). AHVN i a combining architecture between V2V and V2R network. The interoperability between V2V and V2R network i conidered. Then, we developed a Markov chain model in order to determine the MAC of AHVN baed-collaborative. Both theoretical analyi and imulation experiment how that the propoed AHVN trengthen cooperative vehicular ervice and exploit phyical roadway characteritic. Supporting a cooperative ecure beaconing meage communication and developing an efficient packet cheduler in AHVN i our future work. ACKNOWLEDGMENT Thi project wa upported by the Deanhip of Scientific Reearch at Salman bin Abdulaziz Univerity 1433 /ت/ 75 # project (SAU) under the reearch REFERENCES Abbani, N., M. Jomaa, T. Tarhini, H. Artail, W. El-Hajj, Managing Social Network in vehicular network uing trut rule, IEEE Sympoium on Wirele Technology and Application (ISWTA), pp: Alamary, W and W. Zhuang, The Mobility Impact in IEEE p Infratructure le Vehicular Network, Elevier Ad Hoc Network Journal, 10(2): Algana, A., Lin Xiaodong, A. Grami, EVSE: An Efficient Vehicle Social Evaluation Scheme with Location Privacy Preervation for Vehicular Communication, IEEE International Conference on Communication (ICC): 1-5. ANSI-IEEE Standard., 1999.Wirele lan medium acce control (MAC) and phyical layer (PHY) pecification Babu, A.V and V.K. Muhammed Ajeer., Analytical Model for Connectivity of Vehicular Ad Hoc Network in the Preence of Channel Randomne, Wiley, International Journal of Communication Sytem, DOI: /dac Benlimane, A., T. Taleb and R. Sivaraj, Dynamic Clutering-Baed Adaptive Mobile Gateway Management in Integrated V2V-3G Heterogeneou Wirele Network, IEEE Journal on Selected Area in Comm., 29(3): CarTalk, 2000,

15 322 Saied M. Abd El-atty and KontantinoLizo, 2014 Chen, W., S. Cai, Ad hoc peer to peer network architecture for vehicle afety communication, IEEE Communication. Magazine., 43(4): Chun Ng. S, W. Zhang, Y. Zhang, Y. Yang and G. Mao, Analyi of Acce and Connectivity Probabilitie in Vehicular Relay Network, IEEE Journal on Selected Area in Communication, 29(1): Ekram, H., G. Chow, C.M. Leung, R. Mcleod, J. Miic and O. Yang, Vehicular Telematic Over Heterogeneou Wirele network: A Survey, Computer Communication, 33: FleetNet Program, 2001, Lequerica, I., M. García Longaron, P.M. Ruiz, Drive and hare: efficient proviioning of ocial network in vehicular cenario", IEEE Communication Magazine, 48(11): Lequerica, I., P. M. Ruiz and V. Cabrera, Improvement of Vehicular Communication by Uing 3G Capabilitie to Dieminate Control Information. IEEE Net, 24(1): Luan, T., X. Ling and X. (Sherman) Shen., MAC performance Analyi for Vehicle to Infratructure Comm, Proc. IEEE WCNC, 1-5. Murray, T., M. Cojocari and Fu Huirong, Meauring the performance of IEEE p uing n-2 imulator for vehicular network Proc. IEEE EIT, pp: RidongFei, Kun Yang, Xueqi Cheng, A cooperative ocial and vehicular network and it dynamic bandwidth allocation algorithm, IEEE Conference on Computer Communication Workhop (INFOCOM WKSHPS), pp: RITA., Cited-2010., Schonhof, M., A. Keting, M. Treiber, D. Helbinga, Coupled Vehicle and Information Flow: Meage Tranport on a Dynamic Vehicle Network, Elevier, Phyica A, 363(1): Sepulcre, M., J. Gozalvez, J. Harri and H. Hartentein, Contextual Communication Congetion Control for Cooperative Vehicular Network, IEEE Tran. on Wirele Comm, 10(2): Sok-Ian Sou and Tonguz O.K., Enhancing V2V Connectivity through Roadide Unit on Highway IEEE Tran. Veh. Technology, 60(8): Taleb. T and A. Benlimmane., Deign Guideline for a Network Architecture Integrating V2V with 3G & Beyond Network, Proc. IEEE Globecom, pp: 1-5. Wang, C., X. Cheng and D. Laurenon, Vehicle-to-Vehicle Channel Modeling and Meaurement: Recent Advance and Future Challenge, IEEE Comm. Magz., pp: Zeadally, I., R. Hunt, Y. Chen, A. Irwin and A. Haan, Vehicular ad hoc network (V2V): tatu, reult and challenge, Springer Telecommunication Sytem, pp: Zhou, T., H. Sharif, M. Hempel, P. Mahaukhon, W. Wang, and T. Ma, A Novel Adaptive Ditributed Cooperative Relaying MAC Protocol for Vehicular Network, IEEE Journal on Selected Area in Comm, 29(1):