Detection of Packet Dropping Attacks in Wireless Ad-Hoc Network Based On Privacy Preserving Public Auditing

In multi-hop wireless ad-hoc network packet loss can be caused by mainly because of two different reasons. Link error and malicious packet dropping are two different reasons because of which packet loss can be take place. While discovering continuous packet loss in the network. It is hard to find out whether the loss is due to link error or by malicious packet drop. In this paper we can mainly concentrate on insider attack case that is malicious packet dropping where malicious nodes that are part of the route that will drop a small amount of packet which will affect the network performance. Based on traditional algorithm when we compare the packet dropping rate and the channel error rate, the packet dropping rate does not achieve the satisfactory detection accuracy. To improve the detection accuracy, the correlation between lost packets is identified by using the bitmap obtained from individual node. In this paper a homomorphic linear authenticator which is based on public auditing architecture is applied, which allows the detector to find the truthful information about packet loss. The proposed technique is privacy preserving, collusion proof and causes low communication and storage overheads at intermediate nodes. The auditor will collect the information reported by individual nodes and will determine the reason for packet loss by determining correlation between packet loss. Once the malicious node is identified then it is eliminated from route.