Gradient diversity combining for fast frequency-hopped MFSK spread-spectrum communications

The inherent diversity in fast frequency hopped spread spectrum communications systems can provide some protection against interference. With M-ary frequency shift keyed signalling and a diversity L, there are L×M samples from which to make a decision on the symbol transmitted. Many diversity combining techniques have been developed for processing these values. In this paper, new diversity combining techniques are presented which provide robust performance in a changing interference environment. These new combining techniques are based on pattern recognition concepts. The simulation environments used are additive white Gaussian noise (AWGN) and multitone (MT) interference. The simulation results obtained with these new methods are compared to the results obtained using well known combining methods such as linear combining and normalized-envelope detection. It is found that these new combining techniques have excellent performance in MT interference and perform well in AWGN