This paper proposes a quasi-orthogonal correlator-based back propagation neural network (BPNN) pseudo-noise (PN) code acquisition scheme for underwater acoustic direct sequence spread-spectrum (DSSS) communication. The main idea is to use short-length quasi-orthogonal correlators for coarse phase estimation and a BPNN based discriminator for fine phase calculation. When compared with other neuralnet based PN code acquisition schemes, the short quasiorthogonal correlators effectively locate possible phases of the received signal with high tolerance to interference and require no full-length correlation. These capabilities have been verified by extensive computer simulations and the results clearly show that the proposed scheme has the characteristics that are highly suitable for the implementation of underwater acoustic DSSS communication systems.

Included in

Engineering Commons