Research Of Synchronization Techniques In MIMO-OFDM Communication Systems

As the key techniques of broadband mobile communication systems,the combination of Multiple-Input Multiple-Output(MIMO)signal processing with Orthogonal Frequency Division Multiplexing(OFDM)is regarded as a promising solution for combating fading,enhancing the data rates and improving the frequency utilization ratio.In MIMO-OFDM systems,symbol timing offset(STO)and carrier frequency offset(CFO)incur inter-symbol interference(ISI)and inter-carrier interference(ICI),which will seriously deteriorate the performance of the systems.Therefore,proper synchronization algorithms have important theoretical significance and practical value.The demand of efficient synchronization schemes is motivated by following factors.On the one hand,with the increasing signal bandwidth and mobility,the signal distortion stemming from the multipath channel fading will become more serious.On the other hand,with the adoption of the MIMO technique,the enriched dimension of signal space make the received signal suffer from multiple STOs and CFOs.This thesis focuses on the synchronization technique of OFDM and MIMO-OFDM systems,where the main work is organised as follows:Firstly,the synchronization type of digital communication system and the model of OFDM and MIMO-OFDM system are presented in detail.On this basis,the effect of the STOs and CFOs on the OFDM and MIMO-OFDM systems is explored.Simulation analysis of the signal constellation under the influence of STOs and CFOs,highlights the importance of synchronization and provides fundamental base for the following research.Secondly,existing studies for the synchronization in SISO-OFDM system,namely,Moose algorithm,SC algorithm,Minn algorithm and Park algorithm are analyzed.By exploiting the repetitive structures of the CAZAC sequences,a cross-correlation algorithm and an auto-correlation algorithm are further proposed to tackle the foregoing issue.Theoretical evaluation and simulation results demonstrate that compared to conventional schemes,the proposed algorithms can significantly improve the performance of synchronization as well as reduce the computational complexity.Subsequently,a low-complexity synchronization scheme is selected for FPGA implementation according to the practical demands.Simulations and experimental tests demonstrate that the implemented algorithm can support the practical use with satisfactory performance.Thirdly,design principles of synchronization training sequences for MIMO system are given.Under the guidance of these principles,the cross-correlation algorithm and the auto-correlation algorithm proposed for OFDM system are extended to the centralized MIMO-OFDM system.Simulation results show that,compared to the extensively-used EPSP algorithm,the proposed cross-correlation scheme gains boosted efficiency of data transmission while maintains the outstanding synchronization performance.By contrast,the timing performance of the auto-correlation algorithm is not satisfying.Owing to the diversity gain brought by the deployment of multiple antennas,the synchronization algorithms for MIMO-OFDM system primarily outperform their counterparts operating in the OFDM system.Finally,based on a novel synchronization sequence structure,a cross-correlation algorithm is proposed for distributed MIMO-OFDM system.Simulations show that the synchronization performance and transmission efficiency have been improved.Moreover,in order to tackale the compensation of multiple CFOs compensation in distributed MIMO-OFDM system,a scheme based on the zero forcing detection is proposed.Simulation results show that the proposed algorithm can completly compensate frequency offsets and effectively increase the demodulation performance of the receiver.