Study On Channel Estimation Technique Of OFDM Systems

Due to its advantages of high spectrum efficiency, robustness against multi-path fading channel, easy hardware implementation, OFDM(Orthogonal Frequency Division Multiplexing) has been widely used in current high-speed wireless data transmission systems, such as DAB(Digital Audio Broadcasting), DVB(Digital Video Broadcasting) and 802.16 broadband wireless access technologies. It becomes the preferred transmission technique for next generation broadband wireless mobile communication system and broadband wireless transmission of the internet of things.Multi-path effect caused by transmission signal in wireless channel leads to frequency selectivity fading of wireless channel and due to its time-variability under mobile environment. Channel estimation technique is capable of estimating and compensating the channel's impact on signal transmission. It's an indispensable condition for improving the communications quality in mobile wireless channel as well as a key technique for OFDM system.The main contents of the thesis are displayed as follows:1. Wireless channel model has been established by analyzing the signal propagation characteristics in wireless channel. Based on the principle and anti-fading mechanism of OFDM system, we focus on studying pilot aided channel estimation techniques in OFDM system.2. Studying the system performance under different channel conditions, different channel estimations, different interpolation algorithms and different pilot distribution patterns. Simulation results show that MMSE algorithm achieves the best performance, followed by transform domain estimation algorithm, but their implementation complexity is much higher than LS algorithm. Among the interpolation algorithms, cubic spline interpolation performs the best while linear interpolation the worse. The complexity of the algorithm grows rapidly as the improvement of system performance. Different pilot patterns in different channel environments have their own advantages block-type pilot can achieve the best performance and resist longer multi-path delay in slow fading channel. Comb-type pilot can adapt the fast fading channel and is effective in tracking the changing of channel. The irregular distribution of pilots is in the between of the above two pilot arrangement.3. In broadband wireless access IEEE 802.16d Wireless MAN-OFDM system, the traditional channel estimation based on training sequences can not be applied in mobile wireless environment due to lack of real-time tracking ability against channel in one signal frame. According to the situation, the thesis proposes a new channel estimation method with the combination of the preamble training sequences and decision feedback pilots. The algorithm achieves the initial accurate channel estimation by preamble training sequences of wireless frame head. Because of the long interval between pilots in frame structure, the accuracy of channel estimation can be improved by using decision feedback approach. And then the whole channel estimation is accomplished by channel estimation results of preamble training sequences and decision feedback pilots. Compared with the original methods, simulation results show the algorithm can resist frequency selective fading channel with higher doppler shift and improve the performance of 802.16d system in low-speed mobile environment. The channel estimation method proposed in the thesis has been successfully applied in practical system, which is able to operate in complex wireless mobile environment with a pragmatic value.