?
On applying one-sample goodness-of-fit statistics to coded FSK decoding
In the previous works we have considered decoding coded FSK modulation (similar to Kautz-Singleton codes) transmitted in a channel with strong interference by using a two-sample goodness-of-fit statistics. In this work we consider the applicability of different one-sample goodness-of-fit statistics for the decoding. Using a computer simulation we show that using these statistics may yield lower error rates relative to the known decoders. The most notable of them are Anderson Darling statistic that has the lowest error rates for narrow-band interference and Pearson's x2 and Kuiper statistics that are more resilient to wide-band interference. We can conclude that one-sample goodness-of-fit statistics can be used for Kautz-Singleton codes decoding with relatively low error rates in case of strong interference in the channel.