Tow Ship Interference Suppression for Towed Array Sonar via Subspace Reconstruction

As a major interference for towed array sonar, tow ship noise severely decreases the array’s detection ability and effective azimuth range due to its high intensity and wide azimuth spread range. To solve this problem, a subspace reconstruction based technique, which takes advantage of the multiple signal classification (MUSIC) algorithm and tow ship interference’s unique features (including partial coherence, high intensity, and multipath effect), is utilized to suppress the tow ship interference in this paper. This technique is achieved by separating the eigenspace into target subspace, interference subspace, and noise subspace, and then combing the last two into a whole interference-noise subspace to conduct the MUSIC processing. Through principle analysis, we demonstrated that the steering vectors of signals are almost orthogonal to the interference-noise subspace, enabling this technique to suppress the tow ship interference as well as retain the signal output. Moreover, this technique could achieve simultaneous suppression of all the multipath interferences in the interference-noise subspace, making it a superior approach to eliminate the tow ship interference’s azimuth spread. The validity and unique advantages of this method are further demonstrated in simulation and real data processing, which present improved target detectability and display effect.