An adaptive equalizer for digital FM with limiter-discriminator detection in frequency-selective fading channel

Abstract

An adaptive equalizer is proposed for a nonlinear transmission system, i.e., frequency modulation with limiter-discriminator detection. In this method, a linear feedback equalizer at IF band is controlled by evaluating the target function which is defined as mean square error at decision sampling instants. The principle of this method is based on the fact that frequency demodulated signals take given discrete values at the sampling instants if the equalization succeeds. As the adaptive algorithm, the authors adopt the Powell method, that is a general method of search for the minimum value. They propose to repeat the Powell method in order to reach the minimum value avoiding local minimum points. A computer simulation experiment is carried out assuming a two-level FM system with limiter-discriminator detection under two- and three-ray models including minimum- and non-minimum phase shift conditions. The convergence of the equalizer to the minimum value is confirmed. Together with equalization characteristics error rate performance is shown.<>