Research On Adaptive Techniques For SC-FDE Systems In Time Varying Channels

Single carrier with frequency domain equalization (SC-FDE) and orthogonal frequency division multiplexing (OFDM) are both block-wise transmission schemes based on cyclic prefix (CP). They both have immunity to multi-path fading and almost the same complexity and spectrum efficiency. They are two important supportive techniques of physical layer of wideband wireless communications.SC-FDE alleviates peak to average power ratio (PAPR) which is a defect in OFDM systems. Compared with OFDM, the research on SC-FDE emerges later, especially, on adaptive techniques. Recently, the adaptive algorithms based on optimal signal subspace (OSS) have been proposed in SC-FDE. These algorithms are proposed for static channels and can improve the system performance in frequency selective channels remarkably. According to the characteristic in mobile environment, adaptive algorithm of time varying channels is proposed. However, the amount of the feedback information of this algorithm is big, which is not practical.First, we discuss wideband wireless channels. The main contents include multi-path fading channel modeling, WSSUS model about stochastic time-variant linear systems introduced by Bello, and simulation method.Furthermore, we introduce the two block-wise transmissions: SC-FDE and OFDM, and analyze SC-FDE on signal space viewpoint. The concept of optimal signal subspace and the signal transmission based on optimal signal subspace in SC-FDE system are detailed. Later, we introduce adaptive techniques in static channels of OFDM and SC-FDE, and point out the techniques defects in OFDM systems. The OSS dimension determination criterion is also reviewed.Based on adaptive algorithms in static channels and characteristics of time-variant channels, we introduce adaptive algorithm in time-variant channels and point out the amount of the feedback information is big. According to correlation between adjacent frequency sub-channels, the algorithm of reducing the amount of the feedback information is proposed based on channel grouping. The simulation results indicate the algorithm can reduce the amount of the feedback information and do not affect the system's performance.The algorithm of reducing the amount of the feedback information is supported by strong correlation between adjacent frequency sub-channels, which can be bound together as a whole. However, this algorithm can only reduce the amount of the feedback information at a single feedback time, nor reducing the times of the feedback. Then, adaptive bit-loading algorithm is proposed, which utilizes future channel state information acquired by channel prediction and adjusts modulation mode to adapting to varying channel in advance. Therefore, the equalized signal-to-noise ratio (SNR) can be controlled in the needed range, which can reduce channel reselected times and feedback times. The simulation results show this algorithm can improve system performance in low complexity.