Channel Estimation Based On Fft And Improvedalgorithm For Ofdm Systems

With the rapid development of digital communication in recent years, the demand of high data transmission technology is increasing. OFDM technology is paid more and more attention in the wireless environment due to high data rate transmission and the ability to resist inter-symbol interference. OFDM is becoming one of the key technologies of the next generation in wireless communication systems. Coherent demodulation is usually used in OFDM system with the gain of 3dB than differential demodulation, Coherent demodulation requires channel estimation to track the changes of channel response. This paper studies the OFDM Channel Estimation based on FFT and improved algorithm.FFT interpolation is a trade-off between the complexity and performance, which has been widely used in practical application. The two-dimensional channel estimation algorithm between time and frequency domain based on FFT in OFDM system is analyzed first, the performance of algorithm infected by different conditions is compared and studied by using Matlab platform.However, the channel estimation algorithm based on FFT lack of accurate channel information, leading to the problem of insufficient precision in real systems. In this paper, two improved algorithms based on FFT interpolation is proposed, One is the improvement of channel estimation algorithm in pilot location, The other is put the adaptive filtering in channel estimation, the adaptive channel estimation based on LMS After the FFT interpolation is proposed. The LMS algorithm based on MMSE criterion is an optimization algorithm, according to the changing statistics of the input signal, constantly adjust the filter weights in order to achieve optimal output.The results show that the proposed adaptive channel estimation algorithm based on FFT, has strong anti-multipath fading characteristics and the ability to track the fast fading channel quickly, at the same time, because of the low complexity and the easement to implement, the proposed algorithm has a good prospect.