Study and simulation of video communications under wireless environment based on blind source separation

Abstract

A new scheme is proposed to separate several compressed video signals transferred in multiple wireless channels using the blind source separation method. The scheme selects IEEE 802.11b and XVID compression as the wireless communications channel and MPEG-4 video compression, respectively. A simulation model is then made for the video communications under a wireless environment. The model can separate several video signals using blind source separation. Simulations show that the normalized whiten plus cyclic whiten based on cyclostationary (NWCW-CS) algorithm based upon the cyclostationary characteristics of signals has the best separation performance and fast convergence. Besides, the algorithm can solve the mixing of video signals. The image of the transferred video signals decompressed by XVID is nearly consistent with the source ones. The new method meets the requirement of real-time video communications.

Keywords

blind source separation / cyclostationary / video / MPEG-4 / wireless local area network