Blind Estimation On Parameters And PN Sequence Of Non Cooperative Signal

Direct Sequence Spread Spectrum(DSSS) communication has been widely used in secure communication and code division multiple access(CDMA) system in recent decades. Indeed, DSSS transmitters use high rate of pseudo-random(PN) sequence to extend the original baseband signal spectrum. DSSS has a series of advantages, such as large information transmission capacity, good interference immunity and low interception probability. However, these characteristics bring difficulties to communication surveillance and electronic countermeasures. Thus the research on detection and estimation of DSSS signal is of great significance. In this paper, the blind estimation algorithms of DSSS signal parameters and the PN sequence are discussed in the noncooperative communication system. The estimation performance of the algorithms is verified by simulations.Firstly, the conventional DSSS signal model is established, then the correlation characteristic of PN sequence and the spectrum of DSSS signal are analyzed. To realize the blind estimation of the DSSS signal parameters, such as carrier frequency, code period and code rate, carrier double method, auto-correlation method and delay-and-multiply method are studied respectively. In addition, the ensemble averaging technique is applied to further improve the estimation performance. Simulation results show that ensemble averaging can reduce the estimation error efficiently.Secondly, the unknown PN sequence is the principal component of DSSS signal and an important step for blind dispreading. On the basis of the estimated parameters above, the blind estimation methods of PN sequence are analyzed. Eigenanalysis algorithm which has high computational cost is considered at first. Therefore, the variable step-size neural network algorithm, the projection approximation subspace tracking(PAST) algorithm and the maximum likelihood estimation(MLE) algorithm are proposed. In stead of calculating the eigenvalue decomposition, the convergence weights are obtained by iteration, and also the principal eigenvector.Finally, when DSSS baseband signal is synchronous, the principal eigenvector is equal to the estimated PN sequence. When DSSS baseband signal is asynchronous, the PN sequence estimation algorithms combined with double code period window method are proposed. On the basis of the estimated principal eigenvector, sliding window is applied in the eigenvector cutting, and the estimation of PN sequence is fulfilled as well. Simulation results indicate that the improved method can solve the problems in real-time processing and phase ambiguity, which exist in the traditional eigenanalysis algorithm. Meanwhile, the method has good convergence and can realize blind estimation of PN sequence in negative SNR.