Channel Parameter Estimation Algorithms For Ofdm Systems

The growing demand for wireless communications requires reliable and high-rate data transmission over the wireless channel with the arriving of the 4G communications. Orthogonal Frequency Division Multiplexing (OFDM) technique will become the key technology for next generation mobile communication and has widely applied in digital audio and video broadcast due to its high efficiency of frequency utilization and the effective against frequency selective fading, etc. Furthermore, the performance of wireless communication systems is mainly impacted by wireless channel environment, and can be improved by coherent detection, equalization, space-time encoding/decoding technology from using channel information. Therefore, channel parameters estimation is very important, including the channel estimation and the signal-to-noise ratio (SNR) estimation.In order to eliminate inter-symbol interference, the definition of cyclic prefix was introduced in OFDM system, resulting to the cyclostationarity properties of OFDM signal. Based on this characteristic of OFDM signal, the channel estimation and the SNR estimation are studied in this thesis, the main work can be summarized as follows:1,The background is introduced in this section, then the OFDM modulation principle and the cyclostationarity characteristics of OFDM signal are discussed. At last, the multipath fading channel model needed in the data transmission is analyzed.2,Just like most blind channel estimationbased on cyclostationarity characteristics of the OFDM signal, this section propose a novel blind estimation algorithm. Without changing cycle frequency, the new algorithm is formed from using the two related z-transform values of delay parameter. Compared with the existing algorithm, computer simulation results verify that the performance of the proposed One-Cycle algorithm is superior to that of the Two-Cycle algorithm.3,A simplified algorithm based on partial spectrum information can further reduce computational complexity. Based on the proposed One-Cycle algorithm, some key spectrum information instead of the whole spectrum information is used in this simplified method, through analyzing of the cyclic correlation function energy distribution. Compared with the existing algorithm and the One-Cycle algorithm, although the simplied One-Cycle algorithm has some performance loss in high SNR region, its computational complexity is dramatically reduced.4,From analysing the multipath channel, its characteristics, and the cyclostationarity characteristic of OFDM signal, a new algorithm of the noise-to-radio estimation is proposed in this thesis and the theoretical Cramer-Rao Bound result is derived. Computer simulation results verify that the performance of the proposed new algorithm is superor to that of the traditional M₂M₄ algorithm.