Separation Of Coherent Multi-path Signals With PADEC-DEESE Algorithm

The estimation of direction-of-arrival(DOA) of coherent multi-path signals at an antenna array has received considerable attention and it has been one of core techniques of smart antennas. In recent years, development of mobile communication makes people research in algorithms of DOA estimation more deeply, continue to improve existent algorithms and present new algorithms. This thesis is presented based on this background.In this paper, based on through exploring and analysis on the related literatures, the history and state of the arts of wireless direction finding are introduced, the basic knowledge and development trends of wireless direction finding are generalized and summarized, the principle of several wireless direction finding systems, for example, Watson-Watt direction-finding system, interferometer direction-finding system, doppler direction-finding system and so on is showed. After introducing these conventional direction-finding systems, the paper begins to discuss spatialspectrum direction-finding methods that has been very hot from 1990s and further tasks are particulrly put on multiple signal classification(MUSIC) algorithm,space smoothing techiques and estimating signal parameter via rotational invariance techniques. Then DEESE algorithm that Dominic Grenier and Eloi Bosse presented in 1996 and PADEC algorithm that Leila Najjar Atallah and Sylvie Marcos presented in 2004 are introduced. The PADEC algorithm is modified inorder to improve its performance and names the new algorithm PADEC-DEESE. Finally the computer simulated results that includ bias versus SNR, variance versus SNR, bias versus length of data and variance versus length of data are given. The simulated results show that the new algorithm is effective. Compare to MUSIC, PADEC and DEESE, PADEC-DEESE shows better performance while SNR is lower and length of data is short.The novel idea in this paper is that: the decorrelation method of DEESE is applied to PADEC algorithm. The DEESE algorithm has unique decorrelation method that is different from conventional space smoothing method, but the PADEC algorithm's decorrelation method is forward-back space smoothing method. When the number of antenna array's elements is small and the signal-to-noise ratio is low, the DEESE algorithm exhibits excellent performance because of its unique decorrelation method.