The Research On Security Of Routing Protocol In Mobile Ad Hoc Networks

Mobile Ad hoc network (MANET) is a new kind of mobile multi-hops network. It’swidely used because of advantages like flexibility in automatic network building, no need forbase station and so on. However there are also obvious disadvantages about MANET: nodes withlimited resources, topology changing frequently, open radio channel, and the like. Thereforethere are some security problems on it: DoS (Denial of service) attacks, leaks, impersonationattacks, routing protocol security threats. Among them, routing protocol security issues areimportant and difficult, in which detection and defense of the black hole attacks on AODVprotocol is the most representative problem. There are problems on detection of black holeattacks based on AODV: threshold updating mechanism of anomaly detection is imperfect; lackof consideration for the impact of false positive rate on the performance of network.To solve above problems, this paper presents an improved dynamic learning methodIDLAODV to detect and defense black hole attacks based on AODV protocol. It consists of fourimprovement strategies: first, using Hellinger distance to create an anomaly detection model;second, using Chebyshev inequality to set threshold; third, dividing the node into three classes:normal nodes, abnormal nodes, black hole nodes and misjudgment nodes, based on considerationof inaccurate judgment; fourth, separating abnormal node if it is continuously abnormal in thepreviously set observation period, or else allowing its existence.IDLAODV routing protocol was designed and examined in NS2simulation platform, and aseries of comparative analysis performed. The experiment results showed that IDLAODVrouting protocol could reduce the network false positive rate and improve packet delivery ratewhen normalized routing overhead and routing delay was acceptable.