Interference Suppression Study Based On Blind Adaptive Frequency-Shift Filtering

Frequency-shift (FRESH) filter eliminates interferences from signals of interest by using the spectral correlation of cyclostationary signals. As most digital communication signals are cyclostationary, FRESH filtering offers certain advantages for interference suppression in a communication receiver.This paper focuses the research on use of FRESH filtering in satellite communications system to implement co-channel interference suppression in shared frequency communications as well as the related algorithms. The paperdiscusses in detail the core problem in this research -blind adaptive algorithm,proposes three blind adaptive algorithms and analyzes their characteristics respectively by performing computer simulations.Main contents of this paper include:(I)Introduce cyclostationary and spectral correlation concepts of digital signal, analyzes the basic principle that adopts FRESH filtering to suppress interference overlapped with the signal spectrum;(II)Discuss in detail three blind adaptive algorithms: CMA-DD algorithm, LMS-CMA algorithm and non-linear LMS algorithm that is revised ;(III)Analyze and compare through computer simulations with different interference schemes the performances of the above-mentioned blind adaptive algorithms;(IV )The simulation results have demonstrated that among the three algorithms, the general performance of CMA-DD algorithm is the best: for BPSK-modulated signals, when the spectral overlap between SOI and the interference accounts for 45% of the total SOI bandwidth, CMA-DD filter can improve signal-noise ratio (SNR) by 0.7dB compared with common adaptive filter; for QPSK-modulated signals, when the spectral overlap between SOI and the interference accounts for 45% of the total SOI bandwidth, CMA-DD FRESH filter can improve SNR by 0.6dB.