Convergence characteristics analysis on blind separation of DS-CDMA signals with ICA method

Blind estimation of information sequences and pseudo random sequences is of great importance for the security in DS-CDMA system, which remains a hot research topic in reconnaissance and supervision of wireless communication. In DS-CDMA system, the information sequences of different user's are independent and the pseudo random sequences are none correlation, thus independent component analysis (ICA) technology can be adopted to separate the DS-CDMA signals with little prior knowledge. Under non-cooperative situations, the interceptor could not aim at the beginning of the symbol accurately. The offset between the transmitter and the interceptor has great influence on the convergence characteristic of the blind separation algorithm of DS-CDMS signals with ICA method. Certain offsets deteriorate the convergence characteristic obviously. The relationship between the offset and the convergence characteristics is found out, which has great importance in choosing the offset in practice. The reason why certain offsets deteriorate the algorithm is analyzed from the idea of eigen value decomposition. The correctness of theory analysis is proved by the simulation results at last.