Downlink Adaptive Resource Allocation Techniques For Multiuser MIMO-OFDM System

Driven by the increasing popularity of wireless wideband services, future wireless systems will offer kinds of higher data rate applications with very diverse quality of service. However, radio resources such as frequency spectrum would be far from adequate unless advanced technologies are developed to achieve better efficiency of resource utilization. To support such applications under limited radio resources and harsh wireless channel conditions, more and more people start to focus on researching on the new technique MIMO-OFDM for its robust performance over wireless media. To mitigate the frequency selecting and time varying influences of wireless channel, adaptive resource allocation can improve system transmission speed, channel capacity, spectral efficiency and power efficiency. Adaptive MIMO-OFDM system which supports wideband multimedia is one of the most suitable schemes for physical layer.However, there are many problems in adaptive MIMO-OFDM system, for example, the highly computational complexity which caused by the large number of parallel subchannels make it difficult to be utilized in the real system; co-channel interference which increased by the users'spatial relativity reduces the performance of adaptive resource allocation algorithm; and the imperfect channel state information influences the algorithm performance. So the research on the adaptive resource allocation algorithm is very important to the MIMO-OFDM system performance and realization complexity.In this dissertation, adaptive resource allocation algorithms for MIMO-OFDM downlink are studied, which can improve the base station transmission power efficiency, spectral efficiency and reduce the computational complexity. Firstly, the rationales are analysed, for example, co-channel interference cancellation, beamforming (BF), Space-Time Block Codes (STBC) and the combination of STBC and BF (STBC-BF), which are the foundations of the research in following chapters.Aiming at the single-user MIMO-OFDM system, we study the minimum of the system power under Rayleigh block-fading channel while maintaining a constant data rate and target bit error rate, and propose the bit and power allocation algorithm of subchannel space-time-frequency domain dynamic grouping which can reduce the realization complexity. The algorithm decomposes the MIMO-OFDM channels into parallel subchannels by singular value decomposition. All the subchannels are grouped by the space of eigenvalues. In the same space, the subchannels are blocked based on their frequency fading correlations and the blocks are regrouped by the order of their equivalent gains. Then, a Greedy allocation procedure is performed over all the groups. The same allocation is used in same time frame, and an adjusted procedure is performed in different time frame. The proposed algorithm has few performance losses relative to the optimal Greedy algorithm in different data rate and BER, but significantly reduces the computational complexity and the downlink signals which include the transmission parameters of the subchannels.Aiming at the multi-user BF MIMO-OFDM system, we propose the resource allocation algorithm which allocates subcarriers with jointing shared manner and exclusive manner to users according to their spatial correlations under correlated channels. The algorithm improves the performance of the system when the users are highly correlated, and at the same time it also reduces the number of BS antennas required and the computational complexity.In the algorithm the users with high correlation are allocated in the same group. In the same group, subcarriers are allocated in the exclusive manner. While in different groups, subcarriers are allocated in the shared manner. This paper mainly studies the improved null broadening Zero-forcing (ZF) method to mitigate co-channel interference (CCI). The method improves the performance of ZF by broadening the null width based on the null space derivative constraints idea, and reduces the computational complexity by solving null space intersection of user group's channel matrices and using the null steering of the group marginal users.Aiming at multi-user STBC-BF MIMO-OFDM system, with two kinds of channel estimation error at the base station, we investigate the shared manner adaptive resource allocation algorithm which combines STBC and BF under correlated channels. For H type channel estimation error, we propose the jointing Tx-Rx processing zero forcing algorithm, while for the G type channel estimation error, we present null steering algorithm based on main angle of departure. Both algorithms realize CCI suppressing, construct uncorrelated channels for STBC and reduce the number of BS antennas. With the imperfect channel status information at the base station, the adaptive power allocation for the STBC equivalent channels increases the SINR (Signal to Interference and Noise Ratio) at the receiver, and improves the system BER performance. Combining the adaptive bit and power allocation of the subcarriers also make the system performance better.