Study On Wormhole Detection Technology Based Multi-feature In MANET

MANET(Mobile Ad hoc network) is a kind of self-adaptive, self-organized wireless peer-to-peer networks without fixed infrastructure, it was interconnected and organized through the distributed protocol by communication nodes in the system, these features that easy network generation,strong anti-destroying ability and low cost make it widely used in civil and military affairs like battlefield communications,emergency rescue, temporary offices, etc.Despite all the advantages, due to the wireless link, dynamic topology, distributed cooperation,the MANET is extremely vulnerable to various security threats,such as wormhole, DoS (Denial of Service), black hole, grey hole, etc. Among these threats, the wormhole launched by pairs of collusion nodes is a serious attack which disturbs even disrupts MAENT routing algorithms, for instance, the incorrect path selection. Because the wormhole attack is very difficult to be detected, and the way to implement the attack various a lot, the defend against wormhole attacks becomes a very challenging issue. So security methods for resisting the wormhole attack have been an emerging and attractive research community in recent year. In this paper, a new mechanism for detecting the wormhole in MANET is proposed. The related works are outlined as follows:First,the characteristics of MANET routing algorithm and the current research status of wormhole detection technology is analyzed in this paper, the advantages and disadvantages is summarized by comparing various existing wormhole detection scheme. Then, this paper makes a detailed analysis of the features of hidden wormhole and exposed wormhole combining with the characteristics of MANET and AODV routing protocol, and refines a number of unique characteristics of wormhole, including: nodes disappear,the number of neighbors changes abnormally, the number of route hops changes abnormally, being neighbors long time. So based on these characteristics, this paper proposes a novel scheme of wormhole detection based on multi-feature, in this scheme, a super node is introduced to periodically collect neighbor information of each node from whole network, then by synthesizing these neighbor information obtain network topology information and node movement information, and using these information to judge whether it matches the anomaly features of wormhole, after matching multiple features and state transfer, the purpose of accurate and comprehensive wormhole detection is achieved. Finally, a simulation is built in NS2 and the results show that this scheme can effectively and quickly detected hidden and exposed wormhole attack.