FM Noise With Fading Channels and Diversity

TLDR: In this article, the response of an FM discriminator to a fading signal is derived, taking into account the below-threshold noise and the effect of various diversity techniques, including maximal-ratio combining, selection diversity, and equal gain combining.

Abstract: The response of an FM discriminator to a fading signal is derived, taking into account the below-threshold noise and the effect of various diversity techniques. The effect of muting the discriminator output at low signal levels is considered and shown to achieve only marginal improvement. The diversity techniques considered are maximal-ratio combining, selection diversity, and equalgain combining. Only predetection combining is considered for the maximal-ratio and equal-gain techniques. The average SNR at the discriminator output is derived for each form of diversity, assuming that each diversity branch is independent of the others and subject to fading with a Rayleigh envelope distribution. Curves of average SNR plotted against the mean CNR in one branch are presented for various numbers of diversity branches and ratios of IF bandwidth and baseband bandwidth. The curves presented are for maximal-ratio combining only, but it is shown that the results for selection diversity and equal-gain combining may be obtained from these curves by simply applying a correction to the mean CNR. The theoretical results are supported by experimental measurements that show very close agreement with theory.