A Semi-blind Separation Algorithm For The Separation Of Communication Signal And Interference Signal

With the continuous development of the communication, the future battlefieldelectromagnetic environment becomes more and more complex. Because limitedspectrum resources or transmission capacity is limited for the traditional spreadspectrum communication system, it is difficult to further improve its anti-interferenceability. It is an urgent requirement to improve the anti-jamming capability of spreadspectrum communication, since spread spectrum technology is widely used in tacticalcommunication systems. The anti-interference technology of combining Blind sourceseparation with spread spectrum communication is studied in this paper to improve theability of "tolerate interference" and spectrum efficiency for spread spectrumcommunication. The main subject of this paper includes:We propose the semi-blind separation algorithm with prior information for spreadspectrum communication. The spread spectrum communication shared pseudo-randomcode sequence and a symbol rate are the prior information, which can be used in thesemi-blind separation algorithm The priori information can effectively achieve theobjectives of the spread spectrum signal extraction to overcome the uncertainty of thetraditional blind source separation technology. This scheme can improve the separationperformance and reduce the complexity of the algorithm. We also study the relationshipbetween despreading and blind source separation from the view of BER performance.According to analysis and simulation, we find that the traditional blind sourceseparation algorithm can get better performance when the blind separation is performedbefore dispreading, while semi-blind algorithm can get the same performance whendespreading is performed before blind source separation, or not. This analysis providesthe support for further improvement in the anti-jam ability of the system.We combine Independent Component Analysis with Reference (ICA-R) with afixed point algorithm (FastICA) and propose a new semi-blind FastICA algorithm. Withthe prior knowledge, blind source extraction property of ICA-R and less sensitive toGauss noise property of FastICA can be utilized together. Thus, we provide a highefficiency and low complex algorithm for real-time applications. The single channel blind source separation algorithm is proposed based onsemi-blind ICA-R and FastICA algorithm. We realize the effective separation of thesource signal and interference signal with the mixed signal from one antenna. Thismethod can solve the bottleneck of blind source separation with one antenna system. Wecombine the single channel blind source separation with RAKE receiver together. Themethod can improve the anti-interference ability of the system in low SNR andmultipath environments.The blind source separation which is used in spread spectrum communicationsystem, can promote anti-jam ability for military communication system. This work canprovide the supports for the new generation of military communication systems andnetworks.