Node Authentication for Underwater Sensor Networks Based on Time Reversal and LinUCB

Due to the complexity of underwater environment and hash features of underwater acoustic channels, underwater acoustic sensor networks(UASN) has been facing great security challenges. TR process has been generally used for focusing muti-path signals in target detection and communication. In this paper, a node authentication scheme for UASN based on time reversal(TR) process and LinUCB algorithm is presented. This paper proposed, with the hope of authenticating network nodes effectively. Based on network link channel impulse response(CIR), the TR process is used to extract time reversal resonance strength(TRRS) between links. The node authentication problem for UASN is modeled as a contextual multi-armed bandit problem in this paper, where the LinUCB algorithm is utilized to iteratively calculate the decision threshold of TRRS for node authentication. The effectiveness of the proposed online learning node authentication algorithm is verfied through simulation results.