The Study Of Channel Estimation Algorithms In MB-OFDM System

Ultra-wideband wireless communications (UWB) has been considered as one of the main techniques in short-range wireless communications because of its merits such as high transmission rate, good performance, small power, low cost and so on. In ultra-wideband communication systems, to ensure the effectiveness and reliability of communications, the receiver must have an accurate understanding of channel transmission characteristics. Therefore, channel estimation is one of key questions in ultra-wideband communication systems. The main work in this paper is to study the MB-OFDM ultra-wideband system channel estimation algorithms.First of all, this paper introduces the concept of the MB-OFDM ultra-wideband systems, characteristics and so on. Secondly, the paper describes the characteristics of the general wireless channel and the ultra-wideband channel, focusing on the study IEEE 802.15.3a Working Group has put the ultra broadband standard channel models for this channel, and model for the analysis and simulation comparison. Finally, on this basis, focused on the MB-OFDM UWB system channel estimation algorithm. First summarizes a variety of estimation methods, and chiefly analyses the pilot-aided channel estimation algorithm, including the distribution of the pilot, the acquisition of the pilot channel information and all channel information. Then, in ultra-wideband standard channel models, based on Matlab platform under the conditions of MB-OFDM system commonly used by algorithm (LS channel estimate, MMSE DFT and based on the Fourier Transform algorithm domain estimates), as well as their focus on the improvements of the algorithms for the analysis and simulation comparison. Improved to the same algorithms and their respective algorithms from the BER, channel mean square error and other aspects of performance and complexity comparison, the research indicated that the improved algorithms can better track the changes of the channel and improve the overall performance of the system. And then compare horizontally to different algorithms, the research indicated that LS algorithm is the most simple, but not a very good performance; MMSE algorithm for best performance, but the computational complexity is too high, real hard to achieve; based on the DFT transformation territory estimate algorithm, through carries on the time domain filter to the LS estimate result, may effective the performance which enhances the channel to estimate.