Application of the List Viterbi Algorithm for Satellite-based AIS Detection

Satellites receiving Automatic Identification System (AIS) packets in dense areas are particularly prone to AIS channel overload due to the extensive number of vessels. Thus a failure of detection might be caused by the collisions among AIS messages. To improve the detection capability, we propose to exploit the presence of the cyclic redundancy check (CRC) in AIS frames by using the parallel list Viterbi algorithm (PLVA) instead of the classical Viterbi algorithm (VA) often used for decoding AIS signals. The performance of combining the PLVA with AIS post processing including the CRC is studied with two detectors, one coherent and the other differential, in two channel models: a single-user AWGN channel and a more realistic multiple-access AIS channel. We also show the impact of the PLVA parameters on the success recovery rate. The simulation results show that the resulting procedure can significantly improve the packet error rate (PER) at the cost of a limited increase of the computational complexity. The proposed technique could be applied to improve the performance of interference cancellation receivers by significantly lowering the AIS decoding threshold.