CS-Based Homomorphism Encryption and Trust Scheme for Underwater Acoustic Sensor Networks

Underwater acoustic sensor networks (UASNs) have been exploited in many applications. However, due to the complex, unattended and, worse still, hostile deployment environment of the networks, they are vulnerable to many malicious attacks. Presently, researches on how to cope with these threats are extremely restricted due to the limited capability of the sensors. In this paper, we propose a compressive sensing (CS) based homomorphism encryption and trust scheme (CHTS). To identify several malicious attacks such as eavesdropping attacks, compromising attacks, Sybil attacks, wormhole attacks and selective forwarding attacks, etc., we utilize SVM to train trust model and send to each node so that it can determine whether its neighbor nodes are malicious or not. Also, homomorphism encryption is adopted to ensure data confidentiality. Finally, the security analysis shows that the proposed scheme can effectively ensure data confidentiality and identify malicious nodes.