Performance of Post-Demodulator Adaptive Filters for FSK Signals in a Multipath Environment

FSK continues to be an important component of modern communication systems, and discoveries of lower impact methods for mitigating performance degradation due to multipath propagation are needed in many application areas. Previous work suggested the benefit of a simple post-demodulator LMS filter but the focus was narrow and analysis was hindered by the nonlinearity of the demodulation process and the nonstationary signal environment. This paper significantly extends understanding of the post-demodulator filter and demonstrates that the statistical assumptions that align BER performance with MSE performance do not apply in this context, and that simply decreasing the MSE might increase the BER rather than decrease it. An alternative post-demodulator adaptive filter with similar complexity to the LMS filter is proposed which shows improved BER performance. Analysis provides a simplified interpretation of the interaction of system and channel model parameters which may form the basis for more general use of a post-demodulator adaptive filter.