Research On Blind Reconstruction Algorithms Of Spread Spectrum Sequence For Dsss Signals

Spread spectrum communication is the abbreviation of the spread signal spectrum communication.It is a kind of wide band digital communication mode,and its application is very extensive.The direct sequence spread spectrum(DSSS)communication technology is a significant method on spreading spectrum.In this decade,the DSSS communication technology has been applied diffusely in mobile communications,tactical communications,and other aspects.In the case of non-cooperative communication,it is an urgent problem to reconstruct the spread spectrum sequence in a small amout of a priori condition.Therefore,it is a great significance to study the blind reconstruction method of the spread spectrum in low SNR condition.The detection and parameters estimation for DSSS signals was considered in this paper,the center issue is the blind reconstruction of the PN sequence.The main reseatch are as follows:Firstly,the basic composition of the DSSS communication system is briefly introduced,and the mathmatical model of the asynchronous and synchronous DSSS signal is established,and the despread mehod of the DSSS signal is studied.Secondly,the blind synchronization algorithm for asynchronous DSSS signal is studied.The blind synchronization algorithm can solve the problem of phase ambiguity and reconstruct the PN sequence from another angle.Researched the blind estimation algorithms of PN sequence based on eigen value decomposition(EVD).Analyzed the EVD algorithm by the eigen value decomposition of DSSS signals and the principle of EVD algorithm in detail.The improved algorithm is prososed for the splince order of the two fearture vectors in non-synchronous condition.It solved the uncertainty problem when splincing the two main feature vectors.Finaly,the PCA neural network algorithm is studied.The neural network algorithm is analyzed in detail from the aspects of principle and realization method.And this article improved the blind reconstruction method of the spread spectrum codes of the traditional PCA neural network.When achieving the same reconstruction probability,comparing with the traditional algorithm,the performance of the improve algorithm improved by 1dB under the synchronization conditions.When achieving the same reconstruction probability,comparingwith the double periodic neural network algorithm,the performance of the improve algorithm improved by 2dB under the non-synchronization conditions.In this paper,we made the performance comparision between the blind synchronization algorithm,EVD algorithm and PCA neural network algorithm in the sequence correlation,convergence speed and algorithm stability.