Optimization And Implementation For Key Technologies In MIMO-OFDM Transceiver

The demands for mobile transmission performance is ever-increasing according to service's expansion and diversification.Now the third-generation(3G) mobile communication system is technique-matured and meanwhile, the evolution of fourth-generation(4G) is accelerating.The key technology in 4G, such as MIMO and OFDM,get extensive research from the industry.The physical layer of transceiver based on MIMO-OFDM locates at the bottom level of communication protocols. As a carrier of all applications above,the physical layer contains various complicated technology such as modulation,encoding,duplexing and radio module.This thesis aims at optimization and implemention of key technologies in MIMO-OFDM system's physical layer,including frequency-domain interleave,adaptive layer mapping,signal processing for radio and channel estimation in receiver.Research mentioned above is based on the relay transmission system which is the sub-topic of 863 Program "Research,Development and Demonstration for Wireless Communication Technology in High Frequency".First, with the help of simulation platform and channel model,the thesis analyzes system performance requirement and physical layer dataflow,completes the specification of frequency interleave and adaptive layer mapping in transmitter.And second, the high PAPR issue in OFDM system and the nonideal characteristic in RF power amplifier are analysed and a CFR solution is proposed.Finally,for channel estimation,the thesis analyzes the interpolation and time-domain noise suppression in LS design and propose an integrated optimization scheme.The design mentioned is implemented at FPGA platform.In this thesis,combined implementation skill and solution for problems in engineering are introduced.After link-level simulation and hardware testing the frequency interleave can provide link gain of 2dB.Layer mapping is designed to switch adaptively in accordance with data flow.After CFR processing,PAPR of OFDM symbol is inhibited below 7.6dB.Optimized solution for channel estimation can provide processing gain of more 4dB than the traditional LS estimation using linear interpolation.