Study On Blind Recognition Of Space Time Codes

The multiple-input–multiple-output (MIMO) communication system is a key technology fornext generation wireless communication, and the space time code is an important part of the MIMOsystem. It is urgent need to study on blind recognition of space time codes in communicationcountermeasure area, which can provide the basis of countermeasure technology and the technicalsupport for the MIMO system. Blind recognition of space time codes has attracted academicattention, and has important theoretical significance and value of military applications. This thesisstudies on blind recognition of orthogonal space time block codes (OSTBC).Firstly, the basic theory of OSTBC is outlined. Structural characteristics and encoding ofAlamouti codes, the received signal combining, maximum likelihood decoding process areintroduced. Then extended to the general OSTBC, the structural characteristics and encoding forreal orthogonal space time block codes and complex orthogonal space time block codes areintroduced.Secondly, the method for blind recognition of OSTBC based on independent componentanalysis(ICA) is proposed. Assuming that the number of signal transmission time slot and thenumber of symbols contained in each group of transmitted signal are known, ICA technology isused to estimate the virtual channel matrix blindly. Using the orthogonal characteristics of codingmatrix of OSTBC, it is proved that the correlation matrix of the virtual channel matrix in OSTBCcommunication systems is diagonal matrix. And then two characteristic parameters are proposed:the sparsity and the variance of non-diagonal elements of the correlation matrix. A method for blindrecognition of OSTBC using variance of non-diagonal elements (D-ICA), and a method for blindrecognition of OSTBC using variance of non-diagonal elements and sparsity (DS-ICA) are studied.Simulation results demonstrate that the performance of DS-ICA is better than that of D-ICA.Finally, the method for blind recognition of OSTBC based on constant modulusalgorithm(CMA) is proposed. Assuming that the number of signal transmission time slot is known,the minimum description length (MDL) criterion is used to estimate the dimension of the signalsubspace, which is the number of symbols contained in each group of transmitted signals. CMA isused to separate the virtual channel matrix which contains coding matrix information. Because thecorrelation matrix of the virtual channel matrix in OSTBC communication systems is a diagonalmatrix, the parameter of the ratio of the energy of non-diagonal elements of the correlation matrixto the energy of its primary diagonal elements is proposed, and a method for blind recognition ofOSTBC based on the ratio of the energy (EN-CMA) is proposed. And then the proposed DS-ICA, EN-CMA, a method for blind recognition of OSTBC based on ICA and the ratio of energy(EN-ICA), and a method for blind recognition of OSTBC based on CMA and variance ofnon-diagonal elements and sparsity (DS-CMA) are compared with each other. Simulation resultsdemonstrate that EN-CMA achieves best performance.