Comparative Study of Signal Processing Algorithms for Noise Reduction and Sonar Signal Reconstruction

Sonar signal processing plays an important role in extracting information from the unprocessed acoustic data. The processed data ensures accurate target identification. This paper represents a comparative analysis on different signal processing algorithms. The algorithms implemented on the sonar data include Non-negative Matrix Factorization, Least Mean-Squares with variable step size, Hybrid Kalman and Wiener. The different signal parameters such as Signal-to-Noise Ratio, processing time, reconstruction error, convergence speed, noise robustness, and computational complexity for each algorithm are computed. Based on this, scope of each algorithm on different applications are analyzed.