CRVad: Confidential Reasoning and Verification Towards Secure Routing in Ad Hoc Networks

Abstract

Routing security plays an important role in the security of the entire Mobile Ad hoc Network (MANET). In this paper we discuss the issue of private-preserving verification and detection of active attack and passive attack among the peers in MANET. Due to the variability of MANET, attack detection is difficult and usually needs to expose some private data from each node. Attackers can attempt to rewrite, discard the packets or tamper their own log against the expected security requirement to confidentially forwarding the messages. The routers can be easily attacked because of the lack of checking on the message transmission. In this paper, we present CRVad, a method detecting both active and passive attacks in ad hoc network while preserving the privacy of each node. Without introducing any third party to assist this approach, CRVad can be conducted using the information we already know in the MANET and it can detect attacks of multi-hops. CRVad consists of two phases, a reasoning phase with the known confidential information to infer the expected log information of the peers, and a verification phase using Merkle Hash Tree to verify the derived information without revealing any private information of the router. To show our approach can be used to detect the attacks, we conduct our experiment in NS3, and we report the experimental results on the correctness, and efficiency of our approach.