Resource Allocation And Limited Feedback In OFDM System

OFDM is one of the most important candidates in the next generation wireless communication systems. It can effectively combat the frequency selective fading and provide good link reliability. In OFDM related systems the adaptive resource allocation which efficiently utilizes time, frequency and space, improves the system performance has been studied widely. But the base station needs to know the channel information of every subcarrier on the user side before it starts the resource allocation. To FDD systems, downlink channel information need to be estimated by users and fed back to the base station through dedicated feedback channel. When the user number, subcarrier number and the antenna number are big, the feedback information is numerous. So limited feedback is proposed to reduce the feedback information while guarantee the system performance.Limited feedback problems for OFDM related systems are studied in this thesis, and a lot of algorithms are proposed to reduce system feedback overhead for OFDM and MIMO-OFDM systems.We analyze the characteristics of OFDM channel model and the foundation of limited feedback algorithms:the frequency and time correlations of the channel. We also introduce the quantification problem which is unavoidable in digital systems. After analyzing the PDFs of Rayleigh channel, Rice channel and Nakagami channel, we propose a CQI quantification algorithm based on the channel PDF. When the optimization targets are the MMSE of the CQI and the system BER, we respectively derive the quantification formula, get the target threholds through iterative computation, and improve the system performance. Furthermore, we analyze the optimization targets and PDF functions of these quantification algorithms in many practical cases, and point out the usable cases of the algorithms and their directions to be further extended. Compression algorithms have been widely studied by now, and they have also been used in the feedback of wireless communication systems. We introduce a few popular lossless and lossy compression algorithms and their applications in the feedback compression of OFDM systems. A DCT feedback compression algorithm based on sorting is proposed. After the appropriate data block size of it is computed, the proposed algorithm improves the correlation of the feedback CQI in frequency selective channel and reduces the system feedback overhead. In multi-antenna cases, we analyze the correlation among antennas. When there isn't antenna correlation, we analyze the distribution of MIMO matrix eigenvalues. To reduce the CQI feedback information in multiuser MIMO-OFDM systems, an adaptive scheme is proposed according to the correlation among antennas and that of subcarriers in the feedback information, and characteristics of the wavelet transform. Comparing to the two-dimensional DWT-CQI compression algorithm and other two one-dimensional DWT compression algorithms respectively based on antenna correlation and subcarrier correlation, our algorithm has the least feedback overhead. Simulation results show that our compression algorithm greatly reduces the feedback signaling with little BER performance degradation in a V-BLAST+OFDM system, and it can deal with the tradeoff between system performance and feedback overhead well. After analyzing the eigenvalues of the CQI in multiuser MIMO-OFDM systems and deriving their expectations, an improved CQI feedback compression algorithm is proposed which pretreats the feedback information before it is compressed and fed back. This algorithm improves the correlation between the information to be compressed and has good performance. Simulation results show that the proposed algorithm improves the compression ratio of the feedback information and the system throughput performance in a multiuser V-BLAST+OFDM system. It can be effectively used both in DWT and DCT compression algorithms.At last we introduce four kinds of factors which affect the channel performance:channel estimation error, quantification error, delay in the feedback link and the feedback transmission error. We discuss the influence of unperfect channel to the assumed SNR in the base station and the system BER when Max-SNR subcarrier scheduling algorithm is used in the base station.Limited feedback algorithms can effectively save the system feedback overhead, which can be used to transmit more uplink data or more feedback information. The research results and contributions in this thesis can be references for limited feedback in OFDM systems. Some of the proposed limited feedback algorithms are also very meaningful for practice.