The Separated Speech Signals Combined the Hybrid Adaptive Algorithms by Using Power Spectral Density and Total Harmonic Distortion

This paper presents the results of simulation experiments that successfully demonstrate frequency modulation (FM) co-channel speech signals separation by adopting the concepts of power spectral density (PSD) and total harmonic distortion (THD) analyzing adaptive filter combined Fuzzy logic algorithm with the amplitude-locked loop (ALL) separation system in order to improve the signal distortion, co-channel interferences (CCI) of the additive white Gaussian noise (AWGN). The advantage of ALL does not require coding process from modulation, when signals are conveyed by co-channel transmission. In traditional Kalman filter, we find some drawbacks such as low convergence speed and large mean square error (MSE). Hence, we adopted the proposed adaptive algorithm for improving the convergence speed and the MSE value. The improvement simulation results are included separated output signals, the analyses of PSD, THD of the signals, the relation between co-channel signal-to-noise ratio (SNRC), and MSE. It is found the performance of Kalman filter combined Fuzzy algorithm better than other algorithms.