Research On Blind Parameters Estimation Of The Dsss And Soft Spread Spectrum Signals

Direct Sequence Spread Spectrum(DSSS)is widely used in military communications and civil communications due to its strong anti-interference ability,antimultipath fading and low probability of interception.However,it is not suitable to use DSSS signals for information data transmission in communication systems with narrow spectrum resources and high processing gains.The soft spread spectrum signal(or M-ary spread spectrum)derived from the DSSS signal.The soft spread spectrum signal overcomes the inapplicability of the conventional DSSS signal in certain specific conditions due to its unique coding characteristics,so it has attracted much attention.In order to obtain the transmitted information in the non-cooperative communication,it is necessary to know the pseudo-noise(PN)sequence of spread spectrum to perform blind despreading on the intercepted signal.Therefore,the blind estimation of characteristic parameters of DSSS signals and soft spread spectrum signals are of great significance in non-cooperative communication,and the estimation methods of PN and information code sequences have also become the major research contents.In this thesis,the blind estimation of the PN and information code sequence of DSSS and soft spread spectrum signals are studied.The main work includes the following aspects:1.For the problem of the traditional blind despread method for the short code-direct sequence spread spectrum(SC-DSSS)signal,which need to estimate PN sequence to complete the blind dispreading of the SC-DSSS signal.A blind despreading approach based on similarity of SC-DSSS signal is proposed to solve the problem in this thesis.Firstly,the eigenvalue matrix of information code sequence is constructed by calculating similarity value of any two segments of data in the data matrix.Secondly,eigenvalue decomposition(EVD)is applied to the feature information matrix,and the information sequence is estimated by the eigenvector corresponding to the largest eigenvalue.Finally,the PN sequence can be further estimated by multiplying the received signal and the information code sequence.2.For the problem of the traditional single-channel asynchronous short code-direct sequence-code division multiple access(SC-DS-CDMA)signal PN sequence and information code sequence is difficult to estimate,a blind estimation of PN sequence and information code sequence method of SC-DS-CDMA signal in multi-channel synchronous and asynchronous is studied based on parallel factor.Firstly,the signal is modeled as a multi-channel receiving model,based on this point,the SC-DS-CDMA signal of synchronous and asynchronous is segmented according to window length of the single and double PN sequence period,which is constructing the observation data matrix of the signal.Secondly,the observed data matrix is equivalent to a parallel factor model.Finally,the iterative least squares algorithm is applied to decompose the parallel factor,and the information code sequence and PN sequences of SC-DS-CDMA signal is further estimated.3.For the problem of the estimation of the PN sequence period of the multipath soft spread spectrum signal,a blind method based on power spectrum reprocessing was proposed to estimate the PN sequence period of the multipath soft spread spectrum signal.Firstly,calculating the primary power spectrum of the multipath signal.Secondly,the obtained primary power spectrum is used as the input signal to calculate the secondary power spectrum of the signal,and the theoretical analyses show that the peak line of the secondary power spectrum of the signal would appear in the integral multiple of the PN sequence period.Finally,the PN sequence period could be estimated by calculating the space between any adjacent maximum peaks.4.For the problem of the conventional DSSS sequence estimation method is no longer applicable to the soft spread spectrum signal,and an improved K-means algorithm is studied to estimate the PN sequence of the soft spread spectrum signal.Firstly,the similarity measure theory is applied to find out the optimal initial clustering center point of K-means algorithm,and it overcomes the problem that the clustering result is unstable due to the random selection of the initial point.Secondly,clustering results are measured with the absolute value of the average Silhouette Coefficient(SC),and the number of scale of PN sequence can be estimated by the maximum absolute value of the average SC.Finally,the clustering center point corresponding to the clustering number is found to further estimate the PN sequence of the soft spread spectrum signal.