Blind Separation for Wireless Communication Convolutive Mixtures Based on Denoising IVA

In wireless communication systems, signal transmission through a channel can not avoid the influence of noise. When it reaches receiver, is accompanied by time delay and attenuation. Therefore, observed mixture signals at receiver are convolutional mixed with noise contamination. To solve problem traditional frequency-domain based convolutive blind separation methods have poor performance for noise, this paper propose denoise-FastIVA method to separate The basic principle use wavelet transform denoise observation signal, reduce effect on algorithm, enhance robustness fast fixed-point independent vector analysis (FastIVA) algorithm Simulation experiments show effectiveness denoise-FastIVA, under condition that baseband binary phase shift keying (BPSK) frequency (2FSK) modulation 10 bits respectively,the accuracy linear modulization (LFM) has increased from 87% more than 94%; BPSK risen 83% over 97%; 2FSK improved 81% 95%. SNR greater dB, similarity 90%, demixing demodulate completely correctly. Baseband experimental 100 respectively. signal-to-noise ratio 5dB, separated highest coefficient source bit error rate (BER) lowest, compared domain FastIVA algorithm.