Channel modeling and resource allocation in OFDM systems.

摘要

The increasing demand for high data rate in wireless communication systems gives rise to broadband communication systems. The radio channel is plagued by multipath propagation, which causes frequency-selective fading in broadband signals. Orthogonal Frequency-Division Multiplexing (OFDM) is a modulation scheme specifically designed to facilitate high-speed data transmission over frequency-selective fading channels.; The problem of channel modeling in the frequency domain is first investigated for the wideband and ultra wideband wireless channels. The channel is converted into an equivalent discrete channel by uniformly sampling the continuous channel frequency response (CFR), which results in a discrete CFR. A necessary and sufficient condition is established for the existence of parametric models for the discrete CFR. Based on this condition, we provide a justification for the effectiveness of previously reported autoregressive (AR) models in the frequency domain of wideband and ultra wideband channels.; Resource allocation based on channel state information (CSI) is known to be a very powerful method for improving the spectral efficiency of OFDM systems. Bit and power allocation algorithms have been discussed for both static channels, where perfect knowledge of CSI is assumed, and time-varying channels, where the knowledge of CSI is imperfect. In case of static channels, the optimal resource allocation for multiuser OFDM systems has been investigated. Novel algorithms are proposed for subcarrier allocation and bit-power allocation with considerably lower complexity than other schemes in the literature. For time-varying channel, the error in CSI due to channel variation is recognized as the main obstacle for achieving the full potential of resource allocation. Channel prediction is proposed to suppress errors in the CSI and new bit and power allocation schemes incorporating imperfect CSI are presented and their performance is evaluated through simulations.; Finally, a maximum likelihood (ML) receiver for Multiband Keying (MBK) signals is discussed, where MBK is a modulation scheme proposed for ultra wideband systems (UWB). The receiver structure and the associated ML decision rule is derived through analysis. A suboptimal algorithm based on a depth-first tree search is introduced to significantly reduce the computational complexity of the receiver.