Bistable stochastic resonance with linear amplitude response enhanced vector DOA estimation under low SNR conditions

This work demonstrates a superior vector DOA estimation results by linear amplitude response of stochastic resonance with bistable nonlinear model, especially under low SNR conditions. The pre-processing problem of classical intensity based vector DOA estimation method is theoretically analyzed with gain-phase uncertainties, which demonstrate a constraint of linear amplitude response with a certain phase shift for an unbiased estimate of the true azimuth. In this way, linear amplitude response stochastic resonance is parametric modeled with gain-phase constraint, and achieved by the matched stochastic resonance theory with a maximized output SNR and a steady phase lag of π/2. The linear relation between the input and the output amplitude is simulative analyzed under different SNR conditions, which reflect a good linearity with the input amplitude A0 < 1. In contrary to the state-of-art complex acoustic intensity measurement (CAIM) method, a great improvement on estimation performance can be achieved, especially under low SNR conditions. This allows us a new point of view to enhance the vector DOA estimation in the assistance of nonlinear bistable SR effect, and can be a breakthrough innovation guidance for underwater acoustic remote sensing with vector sensors in the future.