Spread Spectrum Sequence Estimation In Long Code Spread Spectrum Signal

Estimation of the spreading codes of the direct sequence spread spectrum (DSSS)signals in non-cooperative communication is an important part of blind dispreading andrestoring the data codes. Its research is an important issue in the field ofcommunications countermeasures. DSSS signals can be divided into short-code DSSSsignals and long-code DSSS signals. Despite a lot of progress for the study ofestimation of the spreading codes has been made in the past, the most of study objectsare the short-code DSSS signals. It is difficult to estimate the long-code signals in thelow SNR case, especially to estimate non-periodic long-code signals. Research resultsof the long-code signals is less than the short-code DSSS signals.This dissertation is a study for the method of period estimating and the spreadingcodes estimating in the non-cooperative communication and low SNR case. The maincontributions of this dissertation include some aspects as follows:1. Three commonly used algorithms in estimating the period of the short-codesignals are studied and compared, and they are expanded to estimating the period of thelong-code signals. The relationship between the three algorithms are analyzed andsimulated. The results show that they are effective in estimating the period of theshort-code signals.2. The eigenvalue decomposition algorithm of the long-code signals is studied andsimulated. To solving the problems of scalar vagueness in the eigenvalue decompositionalgorithm, a method based on shift-sum property of the m sequence is improved andsimulated. The results shows that improved method is better than before. This method isused in estimating non-periodic long-code signals, and is proved that the algorithmused this method is better than other algorithms.3. The principle of triple correlation is studied. The performance of triplecorrelation in white Gaussian noise and data is analyzed. Two algorithms based on triplecorrelation to estimate the spreading codes of non-periodic long-code signals is givenand simulated under different conditions. The eigenvalue decomposition algorithm andtriple correlation algorithm is compared at last.