A Study of Beaconing Mechanism for Vehicle-to-Infrastructure Communications

Transcription

1 Intelligent Vehicular Networking: V2V/V2I Communications and Applications A Study of Beaconing Mechanism for Vehicle-to-Infrastructure Communications Amanda aniel and imitrie C. Popescu epartment of Electrical and Computer Engineering Old ominion University Norfolk, VA {adani023,dpopescu}@odu.edu Stephan Olariu epartment of Computer Science Old ominion University Norfolk, VA olariu@cs.odu.edu Abstract In this paper we study the use of a beaconing mechanism for vehicle-to-infrastructure communication and information exchange in intelligent transportation systems. We provide analytical expressions for the probability of successful connection between roadside infrastructure and passing vehicles showing the dependence on the beacon frequency and distribution of the vehicle speed, as well as on the amount of information to be exchanged between vehicle and roadside infrastructure and the data rate of the wireless link established between the vehicle and roadside infrastructure. The performance of the beaconing mechanism is illustrated for specific scenarios which show that it can be used successfully used in conjunction with wireless systems using the edicated Short ange Communications (SC) standard as well as other types of systems. Index Terms VANET, vehicle-to-infrastructure communications, beaconing. I. INTOUCTION Under present-day technology, traffic monitoring and incident detection on highways employs various technologies such as inductive loop detectors (IL), video detection systems, acoustic tracking systems and microwave radar sensors []. However, many statistics show that over 50% of the installed IL base and 30% of the video detection systems are defective, and the transportation research community is looking for more reliable solutions for traffic monitoring and incident detection [2] [4]. In order to be effective, innovative trafficevent detection systems must take advantage of the most recent advances in technology, among which we note the Vehicular Ad-hoc Networks (VANETs). VANETs employing latest wireless sensor and networking technologies [5] have received increasing attention in the Intelligent Transportation Systems (ITS) community lately. They employ a combination of Vehicle-to-Vehicle (V2V) and Vehicle-to-Infrastructure (V2I) wireless communication and are envisioned to integrate the driving experience into a ubiquitous and pervasive network that will enable novel solutions for traffic monitoring and incident detection [6], [7]. We note that, in V2V systems, each vehicle is responsible for inferring the presence of a traffic-related incident based on reports from other vehicles, which opens the door to various security attacks [8] that could cause vehicles to make incorrect inferences, possibly resulting in increased traffic congestion and a higher chance of severe accidents. Thus, a more meaningful approach to automated detection of traffic-related incidents would be to use traffic monitoring units (TMUs) placed on the side of the roadway to collect information from passing vehicles and to issue traffic advisories informing drivers of potential traffic-related incidents. This would be accomplished through a short-range wireless communication link established between the TMU and passing vehicles over which the information exchange will occur. The potential for predicting traffic-related events using wireless technology is noted in [9] where Bluetooth wireless devices are used for collecting travel time data with the goal of predicting average travel times on freeways. However, the wireless system in [9] is one-way, and only collects information from passing vehicles without providing any traffic-related advisories to the drivers. In this paper we assume that a beaconing mechanism is used by the TMU to inform passing vehicles of its presence in order to establish connection with them and to exchange trafficrelated information and advisories. This is different from [9] where the Bluetooh units are only collecting data from passing vehicles. We note that establishing connection in a controlled manner through the beaconing mechanism is a meaningful alternative to the message broadcast approach commonly used for information exchange in V2V communications [0], which can result in flooding of the TMU with messages from all vehicles that are within its communication range. However, the use of beaconing in a mobile context, where a passing vehicle is within the radio range of the TMU only for a limited time, imposes constraints on the time available for establishing the connection between vehicle and TMU and for successfully exchanging the traffic-related information. Our goal in this paper is to study the probability of successful connection and information exchange between TMU and a given vehicle when beaconing is used in the context of limited TMU communication range. The paper is organized as follows: in Section II we introduce the model for V2I communication, followed by presentation of the beaconing mechanism for establishing the V2I link in Section III. Also in Section III we discuss how the probability of successful information exchange between vehicle and TMU is determined. In Section IV we present numerical examples illustrating the probability of successful information exchange between vehicle and TMU for specific scenarios, and we conclude the paper with final remarks in Section V /2/$ IEEE 746

2 A a b c B d C A B C Fig.. Illustrating a highway segment with several TMUs. Fig. 2. Illustrating the coverage area of a TMU. II. THE V2I SYSTEM MOEL We assume that vehicles traveling on the road are equipped with a tamper-resistant Event ata ecorder (E) such as those mandated by the National Highway Transportation Safety Administration (NHTSA) starting September 200 []. In addition to recording vehicle dynamics and operating parameters as required by NHTSA, the E will also be responsible for recording mobility attributes such as acceleration, deceleration, lane changes and the like which will enable the ITS infrastructure to build statistical databases of long-term traffic-related information based on which it can infer or even anticipate traffic-related events. In addition to the E, vehicles will also be equipped with a short-range wireless communication system used to exchange information with the TMUs embedded in the roadside infrastructure at regular intervals (e.g. every kilometer or so), as illustrated in Figure. The TMUs placed on opposite sides of the roadway constitute an infrastructure unit and are connected by wire under the median. Unlike ILs, adjacent TMUs along the roadway are not connected with each other in order to reduce the overall infrastructure cost. We assume that each TMU contains a GPS device (for time synchronization), a radio transceiver, and an embedded computing device that processes the data collected by passing vehicles to detect traffic-related events. We also assume that the communication range of a TMU shown in Figure 2 is of the order of tens of meters and that it is much smaller than the distance between two consecutive TMUs. In order to communicate with passing vehicles, the TMU radios use multiple access technology to ensure independent coverage of each traffic lane as illustrated in Figure 2. Each lane in the TMU coverage area is assigned non-overlapping sets of orthogonal channels to ensure that there is no interference among the various wireless links established when vehicles in adjacent lanes communicate with the TMU simultaneously. Furthermore, in case more than one vehicle is on the same lane within the coverage area of the TMU, the use of multiple access schemes enables the TMU to establish links with all these vehicles. A wireless link between a vehicle and a TMU is established through a beaconing mechanism when the vehicle enters the coverage area of the TMU: the TMU informs passing vehicles of its presence by transmitting a beacon signal over a standard control channel, and upon receiving the beacon signal from the TMU, the data exchange between vehicle and TMU is initiated. Thus, the success of the information exchange between the passing vehicle and the TMU depends on: the time the vehicle is within the radio range of the TMU, which is a function of the traveling speed of the vehicle; the time required to establish a connection between the passing vehicle and the TMU, which depends on the beacon frequency; the amount of information to be transferred between vehicle and TMU; the data rate of the wireless link established between TMU and vehicle. 747

3 III. THE BEACONING MECHANISM AN POBABILITY OF SUCCESSFUL INFOMATION EXCHANGE To enable connection with passing vehicles, TMUs transmit beacon signals spaced at pre-defined regular intervals of duration T b seconds, such that when a vehicle enters the coverage area of the TMU and receives the beacon signal, a wireless link is established between the vehicle and the TMU, and the data exchange between vehicle and TMU is initiated as shown in Figure 3. Total time t e required for communication between a vehicle and a TMU Time to receive beacon tb Waiting for beacon Vehicle enters TMU coverage area Beacon received from TMU Information exchange time t i ata transmission End of communication between vehicle and TMU Fig. 3. Illustrating the connection setup and information exchange stages for the beaconing mechanism used to enable communication between a TMU and a passing vehicle. Note that the timing diagram in Figure 3 shows the time t b until a vehicle first receives a beacon signal which satisfies the condition 0 t b T b. This is always true under the assumption that the vehicle receives the beacon with probability as soon as it is within the range of the TMU, such that subsequent beacons are no longer relevant to the vehicle and thus, not shown on the timing diagram. Upon successfully receiving the beacon, the information exchange stage is initiated which requires time interval t i, uring the information exchange stage, the vehicle uploads its E data to the TMU and downloads traffic-related information from the TMU. Following the diagram in Figure 3, we note that the distance traveled by the vehicle while waiting to receive the beacon signal is d b and the distance traveled during the information exchange stage is d e. Thus, the total distance traveled by the vehicle while waiting to receive the beacon and exchanging information is d e = d b +d i. We assume that, while a vehicle is within the range of a given TMU, the communication between vehicle and TMU suffers no outages. For short range wireless systems this assumption is usually satisfied, and implies that, in order to successfully exchange information with the TMU, the total distance traveled by the vehicle d e (which is a random variable) should be less than or equal to the size of the coverage area of the TMU. Thus, the probability of successful This assumption implies that the signal power and the corresponding signal-to-interference+noise ratio (SIN) at the receiver are above given thresholds to ensure a specific bit error rate (BE), and that the shift in frequencies due to the oppler effect is tracked and compensated. information exchange between vehicle and TMU is given by = Pr{d e } () and is implied by the cumulative distribution function (CF) of random variable d e. In order to provide an analytical evaluation of this probability we make the following assumptions, which are motivated by the short range of the TMU coverage area: the speed of the vehicle v is constant while it is within the coverage area of the TMU; at any given time we focus on the communication between TMU and an individual vehicle, which is enabled by the use of multiple access schemes. To evaluate the probability (), we focus on a small range of traffic speeds, within which we can assume that vehicle velocities v are uniformly distributed around a given mean velocity v avg. In addition, we assume that the vehicle arrival times are uniformly distributed, so that the time until a vehicle first receives a beacon signal is also uniformly distributed between t b =0and t b = T b. This is a reasonable assumption since a vehicle may enter TMU transceiver range at any point within the beaconing interval with equal probability. Furthermore, the total distance traveled during data transfer between the vehicle and TMU, d e, is the distance traveled over the interval t e = t b + t i as shown in Figure 3, and we consider the worst-case scenario in which the maximum amount of information I max is exchanged between the passing vehicle and the TMU at a data rate, leading to the maximum distance for data exchange d i = v I max With these assumptions we obtain d b = t b v and d e = d b + d i = vt b + v I max = t ev where t e = t b +(I max /) is the total time required for data exchange. Using the probability density functions (PFs) of t e and v we compute the CF of d e as: F e (d e )= f te (t e )f v (v) dt e dv (3) where is the region within the (t e,v) plane in which d e is defined, and t e and v are assumed independent. The six cases defining the region with the associated CF segment are as follows: Case ) /(T b +I max /) </I max <v min <v max for (2) 0 (4) 748

4 Case 2) /(T b +I max /) <v min </I max <v max for T b (v max v min ) [ ( ) ln + I ] maxv min (5) I max v min Case 3) v min </(T b +I max /) </I max <v max for [ ( ) Tb + I max ln T b (v max v min ) I max I max + (T b + I max /) + T b T b + I max / v ] min (6) Case 4) /(T b +I max /) <v min <v max </I max for [ ln T b (v max v min ) ( vmax v min ) I ] max T b Case 5) v min </(T b +I max /) <v max </I max for T b (v max v { [ min ] vmax (T b + I max /) ln I maxv max } + I max T b + I max + T b T b + I max / T bv min Case 6) v min <v max </(T b + I max /) </I max : (7) (8) F b (d e )= (9) Using the CF outlined in these 6 cases and letting be the radio range of the TMU, one can evaluate the probability of successful data exchange for various values of, I max, T b, v avg, and, and some illustrative numerical examples are presented in the following section. IV. NUMEICAL ESULTS AN ISCUSSIONS In order to evaluate the performance of the beaconing mechanism for V2I communications described in the previous sections, we looked at the probability of successful data exchange between vehicle and TMU implied by equations (4)-(9) in specific scenarios corresponding to different beaconing intervals T b, average vehicle speed v avg, data rate over the wireless link, and amount of data I max to be exchanged. We note that, with a value of > 0.8, the TMU can detect potential traffic incidents in about one minute after the incident =250 kbps =500 kbps = Mbps T b in seconds Fig. 4. Probability of successful data transfer as a function of the beaconing interval T b for I max =8kb, uniformly distributed speed with average v avg =70mph, and TMU coverage distance =2m. is first observed/reported [2]. This is a desirable value as it enables rapid notification of drivers by the TMU infrastructure. Figure 4 shows the dependence of on the beaconing interval T b of the system. The average vehicle velocity was chosen to be 70 mph, which is the posted speed limit on many highways in the United States, the maximum information to be exchanged, I max was chosen to be 8kbps, which is more than an order of magnitude higher than the maximum information value chosen in [3]. We note that I max includes not only the traffic incident information that the vehicle and TMU seek to exchange, but also any handshaking data required by the communications protocol for connection setup. In this and subsequent figures, the radio range of the TMU is assumed to be =2m. As expected, decreases with an increase in T b. This is because an increase in the beaconing interval will increase the time needed by the vehicle to receive the a beacon signal upon entering the TMU coverage area, which lengthens the total time required to exchange information and thus, decreases. We note that a value of > 0.8 is achieved for a beaconing interval T b < 0.8 s with the given parameters, and that values are very similar when the data rate of the wireless link is varied from 250 kbps (corresponding to the use of the ZigBee wireless standard) to 500 kbps and Mbps (corresponding to the two version of the SC standard). As expected, an increase in the data transfer rate improves the probability of successful data exchange. From Figure 5, we note that as the average velocity of the vehicles increases, the probability of successful data exchange decreases since the amount of time available for data transfer decreases as speed increases. For this example, a value > 0.8 is achieved for average vehicle velocity values up to approximately 80 mph when T b 700 ms. Figure 6 illustrates the effect of changing the maximum information to be exchanged between a given vehicle and 749

5 5 =250 kbps =500 kbps = Mbps =250 kbps =500 kbps = Mbps v avg in mph Fig. 5. Probability of successful data transfer as a function of the average vehicle velocity v avg for beaconing interval T b = 700 ms, I max =8kb, uniformly distributed speed with average v avg =70mph, and TMU coverage distance =2m. TMU on. From this figure we can conclude that a value > 0.8 is achieved for all values of I max up through I max =25kb when the beaconing interval T b 650 ms. V. CONCLUSIONS In this paper we studied the use of a beaconing mechanism for connection setup in V2I communication systems for ITS, and provided an analysis of the probability of successful data exchange between passing vehicles and TMU placed on the side of the road. Assuming that multiple access technology is used such that a given TMU can communicates with individual vehicles at any given time, we derived closed-form expressions for the probability that successful data exchange occurs within the range of the TMU as a function of the beaconing interval, the average vehicle velocity, maximum information to be exchanged, coverage area of the TMU, and the data rate of the wireless link established between vehicle and TMU. We also illustrated the performance of the beaconing mechanism for typical values of these parameters for communications at highway speeds (up to approximately 80 mph). Our numerical results indicate a probability of successful data transfer between infrastructure and vehicle of greater than 0.8 for a beaconing interval less than approximately 700 ms and maximum information to be exchanged of up 25 kb can be achieved with the SC standard as well as with the ZigBee standard. ACKNOWLEGMENT This work was supported in part by a grant from Old ominion University Office of esearch Summer Experience Enhancing Collaborative esearch (SEEC) program I max in bits x 0 4 Fig. 6. Probability of successful data transfer as a function of I max for beaconing interval T b = 650 ms, uniformly distributed speed with average v avg =70mph, and TMU coverage distance =2m. EFEENCES []. P. oess, E. S. Prassas, and W.. McShane, Traffic Engineering. Erewhon, NC: Pearson Prentice Hall, Third edition, [2] Japanese Ministry of Land, Infrastructure and Transportation, VICS: Vehicle information and communication system, english/, [3] California Partners for Advanced Transit and Highways, PATH, www-path.eecs.berkeley.edu. [4] Virginia epartment of Transportation, Virginia smart road, www. virginiadot.org/projects/constsal-smartrd.asp. [5] O. iva and C. Borcea, The Urbanet evolution: Sensor Power to the People, IEEE Pervasive Computing, vol. 6, no. 2, pp. 4 49, April-June [6] O. iva, T. Nadeem, C. Borcea, and L. Iftode, Context-Aware Migratory Services in Ad-Hoc Networks, IEEE Transactions on Mobile Computing, vol. 6, no. 2, pp , ecember [7] J. Nzounta, N. ajgure, G. Wang, and C. Borcea, VANET outing on City oads Using eal-time Vehicular Traffic Information, IEEE Transactions on Vehicular Technology, vol. 58, no. 7, pp , September [8] A. Aijaz, B. Bochow, F. ötzer, A. Festag, M. Gerlach,. Kroh, and T. Leinmüller, Attacks on inter-vehicle communication systems - an analysis, in Proceedings of the International Workshop on Intelligent Transportation (WIT), Mar [9] M. Martchouk, M. Mannering, and. Bullock, Analysis of Freeway Travel Time Variability Using Bluetooth etection, Journal of Transportation Engineering, vol. 37, no. 0, pp , October 20. [0]. Jiang, V. Taliwal, A. Meier, W. Holfelder, and. Herrtwich, esign of 5.9 GHz SC-Based Vehicular Safety Communication, IEEE Wireless Communications, vol. 3, no. 5, pp , October [] NHTSA, 2006 ruling, NHTSA%20Issues%20Final%20ules%20for%20Automotive% 20Es.pdf. [2]. B. awat,. Treeumnuk,. C. Popescu, M. Abuelela, and S. Olariu, Challenges and Perspectives in the Implementation of NOTICE Architecture for Vehicular Communications, in Proceedings 5 th IEEE International Conference on Mobile Ad Hoc and Sensor Systems MASS 2008, Atlanta, GA, September 2008, pp [3] M. Abuelela, S. Olariu, and M. C. Weigle, NOTICE: An Architecture for Notification of Traffic Incidents, in Proceedings 65 th IEEE Vehicular Technology Conference VTC 08 Spring, Singapore, May 2008, pp