Linear prediction based ETAM method for underwater target bearing estimation

On the basis of conventional extended towed array measurements (ETAM), this paper proposes a new bearing estimation method based on linear prediction techniques called LP-ETAM to estimate the bearings of multiple coherent targets. First, taking advantage of the array linear movement and coherence of underwater signals, the conventional ETAM is used to synthesize the data matrix of extended virtual array from the received data matrix of the physical array. Then, the forward and backward spatial filtering coefficients of linear prediction technique are calculated by the use of the data matrix of extended virtual array, and the aperture of the extended virtual array is increased by extrapolation in each direction. Finally, conventional beamforming technique is adopted to produce the power output which indicates the bearings of multiple targets. Simulations show that compared with the ETAM method, the proposed LP-ETAM method achieves better angular resolution ability and array gain for long-distance underwater targets.