Analysis Of Channel Estimation For OFDM Systems

Channel esimation is not only the key technology of coherent modulation in OFDM,but also the challenged research hot spot in current wireless communication sysytem.In the receiver end,the coherent demodulation and the merge solution for receive siginal both require the information form channel.So the channel stimation research is very meangingful to the wireless communication system. It can't get a good trade-off between performance and computational complexity for conventional channel estimation algorithms. The subject of the paper proposes the new channel estimation principle:adopt dynamic threshold for eliminating noise process in time transform domain.The subject focuses on the research of OFDM channel estimation based on the IEEE802.11a wireless communication standard.Firstly, the subject introduces the research source and background of this project and declares the importance of channel estimation.Then it defines the whole object of the project that it designs a new channel estimation algorithm whose performance and computational complexity are both fine. By the introduction of the principle of OFDM system and the characteristics of OFDM technology, the channel estimation with OFDM systme is describled.Then some other common channel estimation algorithms are introduced. Based on the comparison about their performace and complexity among them, it proposes the linear equalization algorithm for channel estimation,which adds the linear equalization model into the channel estimation. With the valid channel statistics,the key factors,such as the taps parameters and the linear equalization structure, are optimized and designed.Also the error control mechanism is introduced,it can avoid the fault of the linear eatimation that some continous errors exist.Then it creatively proposes the dynamic threshold in transform domain algorithm for channel estimation,which adds the dynamic threshold method into the channel estimation.The new channel estimation not only descreases the channel noise with the dynamic threshold in the time transform domain and gets the channel statistics highly,but also combines with the linear equalization algorithm which can improve the performance of hannel estimation further.The final result can approach the performance of LMMSE but keep the low cpmputational complexity.