Blind estimation of pseudo-random sequences of non-cooperative direct sequence spread spectrum signals in flat-fading channels

Abstract

A novel scheme is proposed to blindly estimate the pseudo-random sequences of direct sequence spread spectrum signals in a non-cooperative context, based on sequential Monte Carlo methodology and eigenvalue analysis technique, with additive white Gaussian noise in flat-fading channels. All the prior knowledge required is only the period of the pseudo-random sequence, which was well developed by the present authors. Because the fading process can be represented by a wavelet basis, the sequential Monte Carlo technique is employed to estimate the transmitted symbols. According to the period of the pseudo-random sequence, the received signals are divided into vectors. The autocorrelation matrix of each vector is accumulated into an estimation matrix, whose corresponding eigenvectors of two larger eigenvalues reconstruct the pseudo-random sequence. Simulation results are provided to demonstrate the good performance of the proposed blind estimation method.