Visible to the public PVad: Privacy-Preserving Verification for Secure Routing in Ad Hoc Networks

TitlePVad: Privacy-Preserving Verification for Secure Routing in Ad Hoc Networks
Publication TypeConference Paper
Year of Publication2017
AuthorsLi, T., Ma, J., Sun, C., Wei, D., Xi, N.
Conference Name2017 International Conference on Networking and Network Applications (NaNA)
Keywordsactive attacks, Ad hoc networks, Cognition, compositionality, data privacy, Data Transmission, diagnostics, distributed cooperation, dynamic topology, expected log data, expected routing paths, MANET, MANETs, Metrics, mobile ad hoc networks, mobile computing, multiple malicious nodes, passive attacks, Peer-to-peer computing, privacy, privacy-preserving verification, pubcrawl, PVad, resilience, Resiliency, Routing, routing discovery phase, routing path, routing security, security, security of data, source node, telecommunication network routing, telecommunication security, verification
Abstract

Routing security has a great importance to the security of Mobile Ad Hoc Networks (MANETs). There are various kinds of attacks when establishing routing path between source and destination. The adversaries attempt to deceive the source node and get the privilege of data transmission. Then they try to launch the malicious behaviors such as passive or active attacks. Due to the characteristics of the MANETs, e.g. dynamic topology, open medium, distributed cooperation, and constrained capability, it is difficult to verify the behavior of nodes and detect malicious nodes without revealing any privacy. In this paper, we present PVad, an approach conducting privacy-preserving verification in the routing discovery phase of MANETs. PVad tries to find the existing communication rules by association rules instead of making the rules. PVad consists of two phases, a reasoning phase deducing the expected log data of the peers, and a verification phase using Merkle Hash Tree to verify the correctness of derived information without revealing any privacy of nodes on expected routing paths. Without deploying any special nodes to assist the verification, PVad can detect multiple malicious nodes by itself. To show our approach can be used to guarantee the security of the MANETs, we conduct our experiments in NS3 as well as the real router environment, and we improved the detection accuracy by 4% on average compared to our former work.

URLhttps://ieeexplore.ieee.org/document/8247105/
DOI10.1109/NaNA.2017.21
Citation Keyli_pvad:_2017