Research On Wormhole Attack Detection And Defense Mechanism In Ad Hoc Networks

Ad Hoc networks are prone to many kind of attacks because of their open network architecture, stringent resource constraints and highly dynamic network topology. Wormhole attack is a severe attack among many attacks in Ad Hoc networks. Wormhole attack mainly focuses on routing protocols. Attackers attract network traffic by giving a shortest path to destination, drop or transmit received packets, then launch black hole attacks or denial of service attacks in future to destroy the availability of networks. As wormhole attack is immune to authentication and cryptography, it is easy to implement. To avoid problems brought by wormhole, a wormhole detection scheme is needed, which can mitigate the impact of wormhole in routing discovery process and make sure the availability of networks.Existing wormhole detection schemes can be divided into four types: based on statistical information wormhole attack detection, based on time anomaly wormhole attack detection, based on distance wormhole attack detection, based on authentication wormhole attack detection. Due to the limited network resources, it will cost high energy and long time to detect the wormhole, which will make great affect on the testing usability. From the wormhole detection predictability and certificate validation perspective presents three classification algorithms: based on fuzzy prediction wormhole detection algorithm(FPWD), trend extrapolation wormhole detection(TEWD), certificate validation classification-ODMRP(CVC-ODMRP). FPWD uses the node to send a packet to determine its surrounding neighbor nodes, then collects a series of location points which stored in contiguous sliding window. According to preliminary statistics neighbor nodes and fuzzy predictive theory, the upper threshold before the node moves to the next position can be predicted. When the node actually moves to the next position, it will detect its neighbor nodes and compares with the upper threshold. When the number of neighbor nodes is greater than the threshold, the point has been attacked by wormhole. TWED is based on FPWD, and uses the fuzzy prediction and sliding window theory to predict the number of neighbor nodes range before the next time. Then we compare it to the actual statistics node number, determines whether it is the wormhole. Under the influence of the wormhole nodes by demand multicast routing protocol to deal with improved performance for Ad Hoc network. To propose a classification based on certificate validation wormhole security defense mechanisms CVC-ODMRP. First use of issuing certificates to detect whether the node has been attacked by wormhole, and the extent of the impact of the wormhole node security grading by neighbors, to provide guidance to select the next communication path and improve the performance of the entire network against wormhole attacks.The simulation results show that the three algorithms have high detection rates and low latency, while achieving the role of wormhole attack prediction and classification of nodes for routing.