Better goodness-of-fit statistics for coded FSK decoding

In previous papers we have considered fast-frequency hopping frequency shift keying. We have proposed a decoder with low error rate in channels with low signal to interference ratio. This decoder uses the Kolmogorov-Smirnov criterion statistics to differentiate transmitted codeword from others. In this paper we consider different goodness-of-fit criterion to improve the decoder error rate. We have discovered that the decoder based on the Barnett-Eison criterion have lower error rates in wideband interference scenario, while the one based on the Anderson criterion is better in case of narrowband interference.