Study Of Channel Estimation Based On Pilot In OFDM Communication Systems

IntroductionOrthogonal frequency division multiplexing is a promising candidate for next generation high-speed wireless multimedia communication systems due to its high data rate, high spectral efficiency, and robust to frequency-selective channels,and it is regarded as the key technology for the systems, which is equivalent to the status of CDMA in 3G presently.When we want to get the benefit form OFDM systems, a series of technologies are needed. One of these technologies, channel estimation, plays an important role in OFDM systems, which is essential to bit and power allocation and signal detection. Without perfect knowledge of channel state information, the OFDM systems either can not work or may incur significant performance loss.Channel estimation methods can be divided into three categories. The first one is the channel estimation method based on pilot or trained sequence. This kind of algorithm has good capability and is easy to realize. The transmission velocity decreases because the introduced pilot symbols or trained sequence occupy some useful bandwidth. The second one is blind channel estimation, which is based on the limited characters of the transmitted information symbols and their statistical trait. This kind of method does not need pilot symbols or trained sequence, so the bandwidth is saved and it can improve the spectrum utilization efficiency. But it has the disadvantages of slow convergence, high complexity. The third one is the semi-blind channel estimation method by using the information from blind channel estimation algorithm and known sampling symbols to finish channel estimation.The improved OFDM channel estimation based on pilotThis thesis mainly focuses on channel estimation in OFDM system based on pilot and training sequence. We analyze 2-D Wiener filter and linear interpolation combined with 1-D wiener filter, and then propose an improved algorithm based on DFT and 1-D wiener filter. The algorithm reduces the complexity comparing to 2-D wiener filter and the estimation performance only slightly degrades pilot channel estimation.Based on MMSE criteria, we know that 2-D Wiener filter is an optimal solution of the linear filter. Using channel's second-order statistics, this algorithm achieves the optimal filter effect through dynamically adjusting filter coefficient. The higher order of 2-D Wiener filter used to estimate the number of pilot, the better performance it will have. In time-varying channel, autocorrelation matrix of pilots in each OFDM signal should be regenerated so as to update filter coefficients. It leads to great arithmetic especially in larger sub-carrier number and higher order of the filter.From the point of view of dimension reduction, two cascaded one-dimension Wiener filter is given on the basis of web pilot channel estimation instead of 2-DWiener filter. With the objective of delay reduction, first one-dimension Wiener filter carries on the OFDM signals including pilots in frequency domain, then filter carries on all OFDM signals to reduce overall delay of channel estimation. Moreover, the design of Wiener filter in actual respect is further discussed. According to different wireless channel power spectrums, we may establish different channel models. The available similar models as mentioned above are: delay and Doppler power spectrum which are subject to the average power spectrum distribution, as well as the delay power spectrum which obeys a negative exponential distribution and the classical Doppler power spectrum. These different models work under different circumstances.Although two cascaded one-dimension Wiener filter has good performance, yet it leads to high computational complexity. In order to further reduce the complexity, linear interpolation is discussed which is used in frequency domain instead of one-dimension Wiener filter. Linear interpolation is a method that interpolates between two adjacent pilots to obtain all data channel estimations. Because of insensitivity to Doppler shift and rapid estimation, it has advantage for easy implemention. However, when the pilot interval is too large, the estimation performance will decline. Especially, in multi-path channel rapid changed in frequency direction, this interpolation method can not dynamically track channels.To improve the performance of linear interpolation, we propose an algorithm based on DFT and one-dimension Wiener filter. For traditional DFT method, zero-padding in time domain is equivalent to interpolation in frequency domain. When the cyclic prefix is greater than the multi-path delay, the energy of channel impulse response concentrates in CP, we may estimate only through the length of CP. Specific process of DFT method is as follows: first, we may divide the received data into several blocks which contain consecutive K samples; second, we may add these corresponding data in each block and obtain the time domain channel estimation. Because of using the receive data in each block, the proposed DFT method is better in performance than linear interpolation which only uses two adjacent pilots in each data channel estimation. Besides,according to different web pilot patterns, we further induce a weighted factor to different locations of pilots. The simulation results show that the proposed DFT algorithm gets obviously improvement than linear interpolation, and effectively reduces complexity compared with 2-D Wiener filter.Conclusions and Future WorkIn this paper, the OFDM channel estimation algorithm based on pilot is studied. We propose an algorithm based on DFT cascaded one-dimension Wiener filter instead of linear interpolation. The simulations clearly show that the performance is better than that of linear interpolation, approximated to 2-D Wiener filter.The channel excess delay is usually assumed to be equal to integral sample, however, this assumption is seldom satisfied in practice. In that case, it will emerge energy leakage. In the future, we have to pay attention to the channel excess delay of non-integral sample time and try to solve this problem in further study.