Hypothesis-feedback equalization for direct-sequence spread-spectrum underwater communications

Direct-sequence code-division multiple-access is considered for underwater acoustic communication networks. Unlike in the majority of spread-spectrum radio systems, intersymbol interference cannot be neglected, and time variability of the channel requires that receiver adaptation be performed at the chip, rather than the bit rate. Adaptive decision-feedback equalization, which has successfully been used for single-user underwater communications, is not directly applicable to spread-spectrum signals because of the delay in the despreading process and the lack of reliable chip decisions. To overcome this problem, a receiver is proposed which feeds back hypothesized, rather than the actual decisions. Numerical examples demonstrate the receiver's ability to cope with time varying channel distortions and preserve the processing gain when conventional, symbol-rate adaptive methods fail.