Algorithm And Application Of Blind Source Separation Based On An Improved Particle Swarm Optimization

Blind source separation is the rapid developed field of research in the late 20th century 80s. It is one of the emerging technologies about the signal processing at present. Blind source separation is the approach to estimate the source signals using only the information of the mixed signals without the information of the source signals and transmission system, without the information of the source signals and transmission system. There are many applications of blind source separation such as wireless communications, image processing, biomedicine, speech processing, radar, geophysical signal processing, fiscal and financial forecasts and so on. Blind source separation has a very high practicability and research value.Independent component analysis is one of the most mature and widely used technology about blind separation of linear instantaneous mixtures. The purpose of independent component analysis is to find a linear transformation matrix. The matrix makes the transformed components independent each other as much as possible. One relates to optimization process of the independence criterion. The traditional optimization algorithm is easy to fall into local optimum and the stability is not good. So it affects the performance of separation. In view of this aspect flaw, this paper proposed a blind source separation algorithm based on the improved particle swarm optimization. The algorithm improved the separation performance.In multi-user detection of DS-CDMA downlink, this paper combined this improved independent component analysis algorithm with subspace MMSE detection. This method can take advantage of both a priori information on the issue and the independence of the user signals. A certain extent, it reduced the bit error rate of the subspace MMSE detection and improve the performance of detector.In multiuser MIMO-OFDM system, the received signals can be expressed as linear mixed-signals in each sub-carrier at the frequency domain after FFT. In this paper, this improved independent component analysis restored the mixed-signals in each sub-carrier. At the same time, the blind channel estimation of MIMO-OFDM system was completed. And there is the order uncertainty and the range uncertainty of independent component analysis. It was resolved by linear pre-coding in transmitter and correlation processing in receiver. The method achieved a good estimation effect.