Interference rejection using time-dependent constant modulus algorithms
Ricardo A. Ramírez-Mendoza, Jeffrey H. Reed, Tien Chang Hsia, B.G. Agee · 1989
Traditional adaptive filters are implemented using timeindependent training-sequence directed techniques. In contrast, the technique prsentcd and evaluated in this paper is based on time-dependent blind weight adaptation. The corresponding adaptive filter is able to exploit properties found in cyclostationary and constant-modulus signals. The steepest-descent CMAs, originally applied to time-independent adaptive processing, arc generalized and applied here to timedependent adaptive filtering for interference rejection. By filtering frcquency-shifted versions of the received signal, CMA-based time-dependent adaptive filters make use of additional spectral information to improve the estimate of the signal-&interest. Cpmputer simulations show the potential of the proposed techntque.