A SECURE ALERT MESSAGING FOR VEHICULAR AD HOC NETWORKS

Transcription

1 INTERNATIONAL JOURNAL OF COMPUTER ENGINEERING & TECHNOLOGY (IJCET) Proceedings of the International Conference on Emerging Trends in Engineering and Management (ICETEM14) ISSN (Print) ISSN (Online) Volume 5, Issue 12, December (2014), pp IAEME: Journal Impact Factor (2014): (Calculated by GISI) IJCET I A E M E A SECURE ALERT MESSAGING FOR VEHICULAR AD HOC NETWORKS Shikha Saju, Smitha Suresh 1 M Tech Student, Computer Science and Engineering, Sree Narayana Gurukulam College of Engineering, India 2 Associate Professor, Computer Science and Engineering, Sree Narayana Gurukulam College of Engineering, India ABSTRACT Safety applications provided by Vehicular Ad Hoc Network is very crucial. The vehicles in the road form a network enabling the passengers with infotainment and security. Emergency messages can be disseminated in the VANET scenario for securing the lives of people by avoiding the situations like chain collisions. The main aim of such messages is to provide safety. So these messages should be genuine one and propagated to all the vehicles in the scenario without any delay. In this paper, clustering technique is used along with the trust relationship to disseminate emergency messages in the network to secure the vehicles from hazardous conditions. Keywords: Chain Collision, Cluster, Emergency Messaging, Intruder, Trust. 1. INTRODUCTION Vehicular ad hoc Network (VANET) is an emerging area with various applications. It is a subset of Mobile ad hoc Network (MANET) consisting of various moving nodes (vehicles) as its members. Vehicles in the network communicate with each other sharing different types of information. The vehicles move at varying speed depending on the nature of the drivers. Vehicles disseminate messages for entertainment and security. Various messages regarding the road conditions, weather, traffic etc can be propagated in the network. Emergency messages play an important role in VANET and hence must be disseminated with great care. The emergency messages are disseminated when critical situations are occurred in the road like accidents, emergency brake or any other hazards. These emergency messages are helpful for the drivers to take necessary steps for avoiding dangerous situations in the road or highway. Every vehicle in the VANET are provided with on board units (OBU) containing various sensors for sensing different network conditions. Since these sensors are of low cost, they can be deployed in the VANET scenario for providing more security for the vehicles in the network and for the lives of passengers in the vehicles. The different on board units communicates with each other sharing much relevant and crucial information helping the drivers to take necessary actions. The network consists of a large number of moving nodes with high speed sharing their information regularly without the help of any road side units (RSU). So the main requirement of the VANET is that safety related messages must be reached to all the member vehicles in the network without any delay. So a major concern should be given to the dissemination of emergency messages (Fig. 1) and the assurance of the reception by all the vehicles. Then only the application of security can be completely implemented in the network assuring lives of the passengers more secured. Along with the assurance of the reception in the messages, the factor of trust must also be considered. Trust relationship must be maintained in the network with all the vehicles and road side units. The main aim of trust management in VANETs is to increase the security, reliability of message or information dissemination, reduction of 1

2 traffic congestions and to ensure the safety of passengers in the network. The trust establishments in the scenario will help to detect any false or wrong information propagated among the members and also to find out the malicious or selfish users. These trustful messages will help the drivers to take appropriate actions during hazards or critical conditions in the road. The main aim in the deployment of vehicular ad hoc network is to enhance the security of vehicles in the road (network). So the trust relationships must be established and evaluated in the real scenario to take accurate decisions during the emergency situations in the highway. The model can manage different vehicles dynamically with the dynamic topology change of network ensuring the security. The untruthful or selfish nodes can be detected easily in the trust evaluation restricting such nodes from further communication inside the network. Thus trust establishment provides more security for the lives of passengers and vehicles in the network. In this paper, the vehicles in the VANET scenario are notified about the emergency event through clustering technique. Also the trust relationship is build to protect from the malicious vehicles. Here privacy of the user is assured with distributed trust establishment. The fake location and time stamp sent by the intruders are detected for more security of the network. The rest of the paper is organized as follows. Section 2 contains the related work. In Section 3, we present the details of our proposed trust management scheme with clustering method and Section 4 consists of the simulation results. Finally, Section 5 includes the conclusion and future work of the work. 2. RELATED WORK 2.1 ROUTING IN VANET Pei et al. [1] discussed about a table driven or proactive routing protocol in which the information regarding every nodes are collected from the neighbouring nodes. Routing table is maintained increasing the storage complexity with the increased network size. Only a partial information is routed for updating, reducing the routing overhead. Perkins et al. [2] discussed the reactive routing protocol which establishes the route on need. This routing protocol is having the sequence number of the destination and hence providing an up to date path to the destination. Excessive memory requirements are not needed and can be used in large scale networks. But the problem is that initial connection setup and communication requires more time. Extra bandwidth is also consumed due to periodic hello messages. Karp et al. [3] proposed greedy perimeter stateless routing which only selects the closest node in the final destination to send the messages. To forward the packet, the node has to remember only one hop neighbour location and these forwarding decisions are made dynamically. The protocol mainly uses the greedy forwarding and perimeter forwarding to disseminate the packets. The information regarding the destination node is never been updated in the packet header of the intermediate node. Santos [4] introduced the cluster based location routing to support the inter vehicle and intra vehicle communications. The clustering scheme used is location specific. The location of the node is very important when the cluster is formed. The route discovery time, end to end delay and number of retransmissions are evaluated. In the case of vehicle to roadside communication the average packet received reduces with the increase in the speed of vehicle. Also the end to end delay, route discovery time and maximum throughput varied very less with the speed of the vehicle. 2.2 TRUST MODELS IN VANET Riaz et al. [5] proposed a robust method to detect false location and time information. The malicious nodes are detected through their messages based on their trust value. The trust value for the genuine nodes will always be greater than the intruders. It achieves security and robustness against the intruders. The problem is that this method will work accurately only if the message is received from the intruder directly. Huang et al. [6] addresses limitations of information cascading and oversampling. This can be avoided by giving weights to the nodes. It means that the nodes who have observed the events are given weight 1 and the intermediary nodes with less weight. So the messages can be received made trustful and be used. Chen et al. [7] designed a framework to support the information sent by the neighbouring nodes and to effectively control the false or malicious information reception in the network. The quality of the message is evaluated in a distributed fashion. The malicious users are prevented from intruding into the network through clustering of the nodes. But the problem is that privacy and robustness are not given much importance in the work. Minhas et al. [8] introduced an approach for modelling trust in VANET environments to restrict the number of reports that are received from other agents. An aggregated feedback mechanism is used which depends on the reliability of the estimated experience-based trust of each other. The confidence value will be calculated and the maximum error rate for the aggregated feedback is noted. The higher rate of confidence states that the message can be trusted in the network. The trust model is aimed to be decentralized, location or time specific, event or task specific and able to support the system level security. The robustness is not all considered in the method which causes problem in the network. 2

3 Sushmita Ruj et al. [9] proposed method to detect the false alert messages and the misbehaving nodes in the network after observing the actions. Here each node has the ability to decide whether the information received is true or false. And this decision is mainly depending on the reliability of recent messages and new alerts with estimated vehicle positions. The method does not rely on any voting schemes and group associations. Along with the false alert message detection, the false location information can also be detected. Gurung et al. [10] proposed the trust model that directly evaluates the authenticity of the messages received from the vehicles. The content of the message is very important during the verification. The road side units or the central servers are not engaged in the checking of trustworthiness of the message. The message classification and information oriented trust model are the two main components of the framework. The two level clustering mechanisms are used to identify the conflict among the information obtained. The trustworthiness of the message received will be calculated in the next component. Merrihan et al. [11] presented a decentralized trust management and evaluation scheme for the different vehicles in the VANET scenario. A combined trust based scheme is used joining the role based and experience based trust. The node is assigned with a category level based on its trust value and the confidence measurement is done based on the degree of trustworthiness of a node s generated reports. A penalty system is introduced to monitor the intruders and trace their actions. It is an efficient message verification mechanism for vehicular communications. It only considers the trust opinions generated from the trusted nodes and discards from the non trusted ones. 3. A SECURE ALERT MESSAGING IN VANET Figure 1: Occurrence of Emergency Event Vehicular ad hoc networks use the safety related applications for providing the passengers safety. There has the chance to occur many hazards in the highway due to many reasons like over speeding or negligence of the driver. But the disastrous conditions are occurred when chain collisions [12] are happened. This condition can be restricted by notifying the vehicles in the highway about the event. The secure alert messaging in VANET scheme mainly alerts the vehicle about the event with trustful messages and avoids the chain collision. The messages are disseminated in the network with clustering technique and the trust values calculated will provide the authenticity of the received messages. 3.1 CLUSTER FORMATION The network can be clustered to disseminate the emergency message faster and securer. The clustering technique will avoid the redundancy of the messages in the network and the broadcast storm problem [13] caused due to the broadcasting method. In the clustering technique, the whole network is divided into clusters or grids and the cluster heads will be elected to transmit the message to the next clusters. These clusters are in charge for disseminating the emergency messages to all the vehicles in the network. So the cluster formation must be faster and dynamic one. The clusters which are formed far away are notified with the help of RSUs Formation of Grid The entire network of the VANET is partitioned into two-dimensional logical grid. The grid cell is a square having a size m*m. Each grid cell is given a grid id for identification. 3

4 Let the VANET scenario be represented by x and y coordinates in X and Y meters and initial length of the grid cell as m. So the two variables p and q can be defines as, (1) The number of grid cells can be found out by using the values obtained in equation (1) and (2), (2) The four corner coordinates must be calculated for the grid cells in the network, (3) The x and y values are calculated using the remainder value and quotient value theorems respectively. The grids are formed in the VANET and each grid cell is given an id to identify them The Midpoint of the Grid Cell is found The midpoint of the grid cell can be identified by using the mathematical equation for finding the midpoint. Here the grid size is known and hence the coordinates of the grid can be found out to obtain the midpoint of the cell. (4) (5) The values of x1, x2, y1 and y2 can be calculated using the following ways, Each node is identified by its grid id and node id. So no other unique names are required for identification. The midpoints are thus calculated Selection of Cluster Head The midpoint of the grid cell is known and the node which is closer to the midpoint is elected as the cluster head. The cluster head can have the ability to disseminate the emergency message to a larger number of vehicles. So the vehicle which lies middle position of the cluster is selected as the cluster head. 3.2 TRUST MANAGEMENT A trust based emergency messaging scheme is used in vehicular ad hoc networks to support trust models used to evaluate the authenticity of information shared among the peers in the network. The topology of aggregation is based on a cluster-based message routing mechanism. For each cluster, a cluster head is chosen randomly and this cluster head has the duty to disseminate the message to the nearby clusters. Trust opinions are aggregated when the message from the (6) (7) (8) (9) (10) 4

5 previous hop clusters and this message is passed on to the cluster members. The cluster head waits for the opinion of its cluster members after disseminating the emergency message in the cluster. Then it is relayed to the next clusters. Trust based message evaluation allows each peer to evaluate the trustworthiness of the message by also taking into account other peers trust opinions about the message and the peer-to-peer trust of these peers. This will help the vehicle to trust the neighboring vehicles. To implement message aggregation scheme a secure and efficient one is vital. One of the solutions to be used is to include signature along with each message sent. Along with the message signatures can be appended, it cannot be maliciously adapted devoid of being detected. Once messages are signed neighbors cannot deny that the messages are sent by them. The signature scheme helps in the authenticity of the received message to protect from non repudiation. The message relay control should be trust based one to sort out the false messages which utilize the insufficient resources such as channel bandwidth to support network scalability. The presence of intruder vehicles must be detected in the VANET, and reduce their node s weight in the network. Thus it controls the harmful situations in the network. Based on the parameters such as location of the message designer and the originality of the message that is the time at which the message was generated are used to detect the wicked messages. It is valuable to detect the trust on data-centric than on the entity-centric. To progress the efficiency of the trust based aggregation method, it will be useful to integrate improvements. Due to multifarious road settings only a division of trust opinions is obtainable for aggregation, thus the usefulness of the system should be calculated in the lack of some the trust opinions. The concentration of vehicles is another parameter for evaluating the aggregation system. Since nodes (vehicles) in VANET have high mobility, thus at some occasion the number of nodes may be very large in number or there may be a lesser amount of vehicles in the road. So the aggregation mechanism used must have the capability to deal with the concentration of vehicles and with data sparse. The emergency event can be detected by the vehicle in the network and the alert messages are propagated in the VANET. The trust relationship is build and the message received is checked for its validity. The malicious vehicles are detected and the false messages received are discarded. The messages from false location also with false timestamp are detected and these vehicles are restricted from injecting further false messages in the network. The functional block diagram is shown in Figure 2. First the VANET topology is created in the network with a number of nodes. Thus the highway is consisting of a number of vehicles moving in opposite directions. The vehicles in the network communicate with each other with periodic beacon messages to update their status. The beacon messages are shared to inform the speed, location etc. of the vehicles moving in the road (network). The occurrence of an emergency event in the highway is detected by the vehicle. Immediately cluster is formed in the road with a small number of vehicles with a cluster head. The cluster head is responsible for disseminating the emergency messages inside the cluster and to other clusters. The middle vehicle in the cluster is elected as the cluster head to propagate the message with more number of vehicles in the group. Then vehicle which receives the message validates the message. After checking the validity of the message, it is disseminated in the network or else it is dropped. Figure 2: Functional Block diagram 5

6 4. SIMULATION RESULTS For the evaluation purposes, the VANET topology was formed in Network Simulator version 2. The transmission range was set to 400m and the simulation carried out in 31 nodes. The main parameters used in NS2 are given in Table I. TABLE I. Channel Type Propagation Model MAC Antenna PARAMETERS USED IN NS2 Channel/Wireless Channel Propagation/Two ray Ground Mac/ Omni directional No. of nodes 31 Transmission range Initially the VANET topology is created in the network simulator with 31 nodes shown in Fig. 3. The vehicles are moving and they are periodically transmitting beacon messages with an interval of 10ms. 400m Figure 3: Initial VANET topology Any emergency event in the network can occur. The occurrence of emergency event in the road causes the dissemination of emergency messages. The clusters are created in the network dynamically to send the alert messages to all the vehicles in the network. These emergency messages are propagated among the clusters which are formed in the network immediately after the occurrence of the event. It is shown in the Fig. 4. Figure 4: Disssemination of Emergency messages through clusters 6

7 The messages received are checked by the vehicles and the false messages sent by the malicious nodes are rejected as shown in Fig. 5. The vehicle receiving the false message verifies the timestamp of the message received and the location of the sender. The false location and timestamp are checked by the receivers and the messages are discarded when they are found to be untrustful without forwarding to the other vehicles in the network. The rest of the cluster members are notified about the intruder in the network by the receiver. 5. CONCLUSION AND FUTURE WORK Figure 5: Discarding the false messages In Vehicular ad hoc Networks, the vehicles are in contact with each other. Each vehicle communicates in regular periods by sending beacon messages. The main difference between other ad hoc networks is that VANET topology changes frequently. Traffic accidents or other emergency situations can occur in the network and the members in the VANET scenario must be notified about the event. So such emergency messages must be reached to all the nodes (vehicles) without any delay. Also, the messages received must not be from an intruder. The authenticity of the messages must be proved to take necessary actions. Here grid based clustering technique is used for the emergency message dissemination. The clustering technique enables the reception of these emergency messages without much delay and the vehicles to take appropriate steps before the occurrence of a hazard. The trust for the messages is provided with the location and time closeness. The verification process verifies all the parameters and hence complete trust is provided. Thus the emergency messages are disseminated very fast without any delay using the clustering technique and the trust is build between the peers in the network. The method is not bulletproof to all types of attacks. So future works lies in establishing a network that is resistant to all types of attacks and protect the VANET from all types of intruders. REFERENCES [1] Pei G., Gerla M. and Chen, T.-W (2000), Fisheye State Routing: A Routing Scheme for Ad Hoc Wireless Networks, Proc. ICC 2000, New Orleans, LA, June [2] Perkins C., Belding-Royer E., Das S. (July 2003) Ad Hoc On-Demand Distance Vector (AODV) Routing. [3] Karp B. and Kung H. T. (2000), GPSR: Greedy Perimeter Stateless Routing for Wireless Networks in Mobile Computing and Networking, pages , [4] R. A. Santos Supporting Inter-Vehicular and Vehicle-Roadside Communications over a Cluster-Based Wireless Ad Hoc Routing Algorithm. [5] Riaz Ahmed Shaikh, Ahmed Saeed Alzahrani, Intrusion-aware trust model for vehicular ad hoc networks. Published in Wiley Online Library in Security and Communication Networks (Aug 2013). [6] Zhen Huang, Sushmitha Ruj, Marcos Cavenaghi, Amiya Nayak. Limitations of Trust Management Schemes in VANET and Countermeasures, In IEEE 22 nd International Symposium on Personal Indoor and Mobile Radio Communications (2011). [7] Jie Zhang, Chen Chen, Robin Cohen, Trust Modeling for message Relay Control action decision making in VANETs. Published in Securiy and Communication Networks, volume 6 (2013). 7

8 [8] Minhas U. F., Zhang J., Tran T., Cohen R. Intelligent Agents in Mobile Vehicular Ad Hoc Networks: Leverging Trust Modeling Based on Direct Incentives for Honesty. In proceedings of the IEEE/WIC/ACM International Conference on Intelligent Agent Technology, Toronto, Canada, vol.2, pp (2010). [9] Sushmita Ruj, Macros A. Cavenaghi, Zhen Huang, Amiya Nayak and Ivan Stojmenovic. On Data-Centric Misbehavior Detection in VANETs. Published in Vehicular Technology Conference (VTC Fall), 2011 IEEE, 5-8(2011). [10] Sashi Gurung, Dan Lin, Anna Aquicciarini, and Elisa Bertino, Information- oriented Trustworthiness Evaluation in Vehicular Ad-hoc Networks. Published in the 7 th International Conference on Network and System Security (2013). [11] Merrihan Monir, Ayman Abdel-Hamid and Mohammed Abd El Aziz. A Categorized Trust-Based Message Reporting Scheme for VANETs. Advances in Security of Information and Communication Networks in Computer and Information Science vol. 381, Springer Link (2013). [12] Q. Xu, R. Sengupta and D. Jiang, Design and Analysis of Highway safety Communication protocol in 5.9 GHz dedicated short range Communication Spectrum, IEEE Trans. Veh. Technol., vol. 57, no. 4, pp , [13] O. K. Tonguz, N. Wisitpongphan, J. S. Parikh, F. Bhai, P. Mudalige and V. K. Sadekar, On the broadcast storm problem in ad hoc Wireless Networks in Proceedings of International Conference on Broadband Comunications, Networks and Systems, 2006, pp [14] Dr.S.Rajalakshmi and S.T.Tharani, Signature Seeking Drive for Incentive Based Ad Dissemination in Vehicular Networks, International Journal of Computer Engineering & Technology (IJCET), Volume 4, Issue 3, 2013, pp , ISSN Print: , ISSN Online: [15] Mitul K. Patel, Study of Localization Techniques in Vehicular Ad-Hoc Networks, International Journal of Computer Engineering & Technology (IJCET), Volume 4, Issue 4, 2013, pp , ISSN Print: , ISSN Online: