NEURAL-NET-BASED RECEIVER STRUCTURES FOR SINGLE-AMPLITUDE AND MULTIAMPLITUDE BAND-LIMITED SIGNALS IN CCI AND ACI CHANNELS

This paper presents analysis and performance-evaluation results for se veral neural-network-based nondecision-feedback receiver structures, w hich improve the performance of bandlimited single- and multiamplitude signals transmitted over additive interference channels, such as coch annel interference (CCI) and adjacent channel interference (ACI). In p articular, we propose, analyze, and evaluate a training algorithm for Ngquist-filtered single- and multiamplitude signals, based upon a nove l nonuniform signal-sampling technique. We also introduce a novel nonl inear activation function for multiamplitude signals and evaluate its performance via computer simulation and in conjunction with various ba ndlimited signaling formats, detection techniques, and neural-network structures. Bit-error rate (BER) performance-evaluation results of the proposed neural-network receivers for coherent and noncoherent detect ion of Nyquist- and Butterworth-filtered single- and multiamplitude si gnals have shown performance improvements in the presence of CCI and A CI.