Blind Estimation Of Multiple Pseudo-Random Codes Of Long-Code Direct Spread Spectrum Signals

The problem of pseudo-random(PN)codes estimation of spread spectrum signals in non-cooperative communication systems has been widely studied by scholars from all over the world.Because periodic long and short codes direct sequence spread spectrum(PLSC-DSSS)signal,non-periodic long code direct sequence spread spectrum(NPLC-DSSS)signal,non-periodic long and short codes direct sequence spread spectrum(NPLSC-DSSS)signal and periodic long and short codes direct sequence code division multiple access(PLSC-DS-CDMA)signal have the complex structure,the PN codes estimation of these signals which use a pair of PN codes serving as spreading code and scrambling code is very difficult.This dissertation mainly studies the blind estimation of PN codes in non-cooperative communication systems.First of all,the basic principles and characteristics of DSSS and CDMA communication systems are discussed,and the classification of DSSS signals is given.The signal models of short code direct sequence spread spectrum(SC-DSSS)signal,the long code direct sequence spread spectrum(LC-DSSS)signal and the LSC-DSSS signal are given.The current researches on PN code estimation of DSSS signals in non-cooperative communication system are also introduced in detail.Secondly,the blind estimation of PN codes in periodic long and short codes direct sequence spread spectrum signal based on eigenvalue decomposition and triple correlation is studied.With the further study of triple correlation function(TCF)of m-sequence,the relationships between forward and backward TCF peaks are deduced.Based on the above theoretical results,the blind estimation algorithm of long and short PN codes using the eigenvalue decomposition,the unitary matrix method of eliminating fuzzy relations and triple correlation is put forward.The simulations prove that this algorithm can estimate the long and short PN codes of PLSC-DSSS validly,and outperforms the existing algorithm.Thirdly,the PN codes estimation of non-periodic direct sequence spread spectrum signals is studied.(1)The PN code estimation of NPLC-DSSS signal based on goodness of fit test is researched.The binary hypothesis test model and probability distribution function of TCF peak detection of NPLC-DSSS signal are derived in detail,and the TCF peaks can be detected precisely by the methods of cycle removing mean and goodness of fit test.Based on the above methods,the long PN code of NPLC-DSSS signal can be estimated.The simulations prove that the goodness of fit test method can raise the accuracy of PN code estimation.(2)The PN codes estimation of NPLSC-DSSS signal based on forward and backward TCF peaks is studied.Aiming at NPLSC-DSSS signal which contains long scrambling code and short spreading code,the common peaks of long and short PN codes are detected by the forward and backward TCF peaks properties of m sequence,and then the algorithm of the PN codes estimation based on the common peaks is proposed.Simulations show the proposed algorithm has well performance.Finally,the blind estimation of PN codes in PLSC-DS-CDMA signal based on matrix completion(MC)and triple correlation is studied.The PLSC-DS-CDMA signal is represented as a matrix model with missing data for multi-user short code one in theory,and the composite code matrix estimation is modeled as a blind source separation problem.Matrix completion theory is used to estimate the composite code subspace,and a method of the composite code sequences estimation is proposed based on the singular value thresholding algorithm and Fast-ICA algorithm.Then the delayed triple correlation algorithm is used to estimate the long and short PN codes form the composite code of every user.The simulations show that the algorithm can achieve well performance of PN codes estimation in PLSC-DS-CDMA signal.