Spreading Code Estimation Using Inherent Linear Relation in Non-Cooperative Contexts

Abstract

In non-cooperative contexts, estimating the parameters of spreading code generated from linear feedback shift registers (LFSRs) is essential for recovering messages from received direct sequence spread spectrum signals. This article proposes a method to improve estimation performance of spreading codes by leveraging the inherent linear relation in spreading code due to LFSR. To achieve this, the spreading code is initially estimated by a conventional algorithm. Subsequences are then generated by decimating the initially estimated code, and their generating polynomials are identified using the Berlekamp–Massey algorithm. Using the polynomials, sequences satisfying the linear relation are reconstructed, and the original spreading code is recovered as the reverse decimated sequence with the smallest Hamming distance to the initially estimated spreading code. We validate the method through an analysis of estimation accuracy improvement and computational complexity reduction through simulations.