| Compressed sensing has properties of high compressibility and good recoveryperformance when processing sparse signals. It’s widely used in many fields such assignal processing, image processing. In recent years, the compressed sensingtechnology is used for signal estimation and detection in wireless communication.Such algorithms can obtain good estimation and detection performance. Also, thesealgorithms can simplify pilots to improve the transmission efficiency. Compressedsensing has gradually become a hot research direction in wireless communication.In the area of wireless communication, most of the traditional detection methodsuse the channel information to implement the detection process. The pilots are used forchannel information estimation in conventional algorithms such as least squares (LS)method, minimum mean square error (MMSE) method, etc. We also make use of pilotsto estimate noise variance and the system’s signal to noise ratio (SNR). At thereceiving end, channel information is used for equalization; including the maximumlikelihood (ML) detection, zero forcing (ZF) detection, MMSE detection methods.With the development of wireless communication technology, orthogonalfrequency division multiplexing (OFDM), multiple input multiple output (MIMO)technology have been widely used, wherein the spatial modulation (SM) technology isproposed as a new kind of MIMO technique. By extending the dimension of theantenna, it can greatly increase the transfer efficiency. With the aid of these newtechnologies, information process is extended to time, frequency, and space domain,making its estimation and detection more complex. Also, with the dimension extension,the signal to be processed in some dimension is sparse.In the conventional estimation methods, a large number of pilots are required toimplement channel estimation, which reduces the spectral efficiency of the system. TheMMSE algorithm is possible to achieve good estimation performance, however, itneeds a priori information of the channel and it has a quite high complexity. In the caseof spatial modulation, the traditional ML detector has good detection performance, but the complexity is too high, the ZF detection and MMSE detection performance are notsatisfactory. Therefore, this article attempts to exploit compressed sensing methods tocarry out OFDM channel estimation and noise variance estimation, and signaldetection in spatial modulation, and performance assessment is also analyzed. Specificresearch contents are as follows:Chapter1gives a brief introduction of estimation and detection problems as wellas compressed sensing technology, and indicates the paper content arrangements.Chapter2analyzes the compressed sensing model, presents the development ofcompressed sensing theory, and describes the system framework. Signal’s sparserepresentation, the observation matrix design are also introduced, simulation results areanalyzed based on a variety of reconstruction algorithm.Chapter3gives an introduction of channel characters and OFDM system.Common channel estimation algorithm and noise variance estimation algorithm arebriefly compared. Compression sensing based channel estimation is introduced. What’smore, according to sparsity adaptive matching pursuit (SAMP) algorithm, a new jointestimation algorithm is proposed.Chapter4analyzes the spatial modulation system model and its derivative of thegeneral spatial modulation (GSM) and generalized space shift keying (GSSK) model.Based on the above model, linear equalizer detection algorithm, ML detectionalgorithm and compressed sensing based detection algorithm are introduced; what’smore, some improvements are made. We deduce that, in GSM and GSSK model, usingcompressed sensing technology for detection can significantly reduce the complexityof the system, and obtain a good detection performance. Simulations are carried out toverify the performance of the proposed algorithm.Chapter5summarizes the research work, gives a brief discussion about what needto be improved in the future. |
PDF Full Text Request