Algorithm of Bispectral Video Signal Recovery under Conditions of Industrial Noise

摘要

A binary communication channel under the action of additive Gaussian and non-Gaussian broadband and narrow-band interferences is considered in order to identify the possibilities of using the method of bispectral analysis for the detection and recovery of the useful signal. The discrete random process with SL Johnson distribution is taken as non-Gaussian noise. The estimation of periodic signal detection and recovery effectiveness has been held by means of bispectral processing the mixture of noise and a priori unknown signal. The digital simulation was implemented in MatLab software. It has been shown the method of bispectral analysis allows recovering the signal more qualitative in comparison with the ordinary spectral analysis when SNR is close to unity. The use of developed algorithm allows detecting the periodic pulse packets at large SNR values.