Research And Implementation Of Channel Estimation In Multi-antenna OFDMA Systems

Space-time coding utilized in multi-antenna OFDMA systems can not only meet the transmission requirements of high speed and large capacity but also enhance stability of communication systems. The algorithms of space-time decoding which are commonly used in systems depend on the accurate channel state information to a great extent, so channel estimation is one of the key technologies which affect multi-antenna OFDMA system performance.The channel estimation algorithms and implementation in multi-antenna OFDMA systems are mainly studied in this thesis. Firstly, OFDMA and MIMO as well as channel estimation principles are introduced. Secondly, the model of multi-antenna OFDMA system is designed which can afford simulation platform to test the channel estimation algorithms. Then the channel estimation algorithms based on the multi-antenna pilot pattern are theoretically researched. Combining the design of virtual pilot and two-dimensional polynomial fitting algorithms, the Virtual Pilot Algorithm (VPA) is proposed, which is fit for fast moving scenes. The VPA can use the relativity of the two-dimensional channel information to fit the channel, so the performance of channel estimation can be improved. The simulation results show that the mean square error performance and bit error rate performance of VPA are better than others in fast moving scenes. Finally, the quantization schemes of multi-antenna channel estimation and space-time decoding in PB3(ITU Pedestrian B, Speed 3km/h) channel model are given, and FPGA implementation is completed.