Research On Adaptive Technology Based On Orthogonal Frequency Division Multiplexing

The multipath fading propagation channel is occurred in Wireless communication, which can be overcome by dividing the broad bandwidth into some sub-channels with Orthogonal Frequency Division Multiplexing (OFDM). As an effective high-speed digital transmission mode, OFDM and the adaptive transmission have been comprehensively considered as the major candidate technique for future mobile communication systems. The adaptive technique based on OFDM has been widely applied in many fields and accepted by many international communication standards. Our work is supported by the National High Technology Research and Development Program of China under Grant No.2001AA123014 and the National Science Foundation of China under Grant No. 603905405.Firstly, the wireless channel characteristic and the OFDM model are analyzed. The system performances of OFDM can be significantly improved by adaptive technique. The adaptive modulating strategy based on the minimal transmitting power criterion is structured. Assuming the transmitter and receiving ends have the perfect channel state information (CSI), the genetic algorithm to deal with the adaptive power allocation and bit loading under multi-user environment was present. It was intuitionist and brief by the Matrix coding,the local optimum was avoided and the global optimum solution was obtained by the crossover and mutation operation. And the characteristic of parallel computing saves running time and improves the calculating efficiency.Secondly, we research the application of adaptive OFDM technique in MIMO-OFDM system. It is well know that the MIMO can improve the system capacity and wireless links'reliability because of the diversity utilized in the space and time region and the use of multi-antenna array in transmission and receiving ends. Unfortunately, the transmitter usually couldn't achieve the perfect CSI in practical MIMO-OFDM system. Therefore, the adaptive MIMO-OFDM based on the mean feedback partial CSI was focused. Assumed the receiver has the perfect CSI and the transmitter has mean feedback CSI, the scheme of concatenating TCM, 2D eigen-beamforming with OSTBC was put forward and the correspondent bit error ratio was deduced. The improvement come from the TCM coding gain, the dynamic sub-carrier allocation and the multi-antenna array gain in the suggested system.Thirdly, the adaptive transmit scheme based on the finite rate feedback is analyzed in MIMO-OFDM system. Assumed the receiver has the perfect CSI while the transmitter has the limited CSI, the receiver only sends the index of the best beamforming vector, which is selected in a predetermined codebook, to the transmitter through a low-rare, errorless and zero delay feedback link. The Grassmannian line package can be used to design the quantized beamforming codebook based on the optimization criteria of maximum signal-to-noise ratio (SNR). The optimization beamforming codebook based on maximum ratio transmit (MRT), equal gain transmit (EGT) and the selection diversity transmit (SDT) are analyzed by the Grassmannian line package respectively. And the quantized distortion is given. The low bounds on the codebook size are derived under the given capacity or SNR loss.Finally, the adaptive OFDM technique is employed by the UWB system. Three kinds of power allocating schemes are presented based on the optimization criteria of maximizing capacity with different tradeoff among performance, complexity and feedback bandwidth overhead. (1) Assumed the transmitter has the perfect CSI, the classical water filling scheme is present to deal with the power allocation of each sub-carrier. (2) The grouped power load is present, which is solved by KKT method. All the subcarriers are grouped, and the unique power is loaded on every group instead of every sub-carrier, so the complexity is lowered with performance penalty. (3) Assumed the transmitter just have every group's average attenuation factor, the limited grouped feedback is present. This method reduces the calculation complexity and the channel feedback overhead, and improve the system resources utilization.