| Orthogonal Frequency Division Multiplexing (OFDM) technology is widely used in various wireless communication systems including Digital Audio Broadcasting (DAB), Terrestrial Digital Audio Broadcasting (DAB-T), and is a core part of the Next Generation mobile communication system standards. OFDM system, usually using pilot aided channel estimation, firstly estimate the pilot position channel frequency response, and then use some interpolation algorithms to estimate other channel frequency response on other locations. But such methods do not take into account the sparsity of the channel, such as the underwater sound channel, high-definition television (HDTV) channel, of which the channel impulse response of only very few positions is not zero and that of the remaining positions are zero[1]. The channel's sparsity can be exploited to reduce the number of pilots for the sake of bandwidth efficiency. This paper studies how to use this small amount of pilot subcarriers to estimate the channel, i.e., design the interpolation algorithm which calculates the channel frequency response of the subcarriers other than the pilots.Recent years, foreign counterparts have proposed to use compressed sensing theory(Compressed Sensing, CS) to estimate the sparse channel, dramatically reducing the number[2] of pilots required. Compressed sensing is the latest research area in signal processing community, allowing the effective reconstruction of sparse signals from a very limited number of samples[3][4]. This thesis proposes the scheme of applications of compressed sensing technology on channel estimation in OFDM system, also give a test to the improved iterative minimum norm method (also called iterative basis pursuit method, RBP)[5], and in the last section, it compares with the 11 minimum norm method which is widely used (also called Basis Pursuit Method, BP)[2] [6].Simulation results show that, RBP is more accurate in channel estimation, and thus can significantly reduce bit error ratio (BER). |
PDF Full Text Request