Siganl Detection Algorithm Research And Implementation In High-Speed Mobile Communication System

When data is transmitted through band-limited channels in mobile communications, the inter-symbol interference (ISI) caused by channel distortion and multi-path propagation will make decoding error. To remove this kind of interference and achieve as high transmit rate as possible, signal detection techniques, which include mainly equalization techniques, are necessary here to compensate signal distortions and improve the performance of the receiver.In this thesis firstly, the origin of ISI is analyzed and the performance and regular model for mobile channel is provided, also, various equalization techniques for mobile communication channel are introduced here. Based upon the above, the main content of this thesis contains the realization and optimizing for space-time detector based on single-point DFT-domain equalization, and the fix-point simulation and hardware realization for Iterative space-time detector based on interference cancellation. The algorithms and realization structures for both detectors are described here.Space-time detector based on single-point DFT-domain equalization is derived from the minimum mean square error (MMSE) criterion, and Space-time Joint and DFT-transform are the two main sub-modules of DFT detector. Based on time-division pipeline techniques, only one complex-multiplicator and one FFT-transform module and the corresponding control circuit are employed to realize the detector and meet the system requirement in detection capability and hardware occupation. Considering the performance limit and un-reasonable source occupation caused by the FFT core generated in ISE in the realization scheme, the more suitable FFT-transform module is compiled to improve the performance and to optimize the hardware occupation structure.Iterative space-time detector based on interference cancellation is an iterative detection scheme, in which firstly the interference signals (calculated from the initialization of the mean value and variance of signal) is subtracted from the received signal, then the estimate of the signal and the noise variance can be calculated out, and subsequently we get the condition-probability for the transmitted signals and the mean value and variance of the estimate signal are renewed, which in turn will be substituted into the next iterative operation. The final estimate values for the signal and its variance after several iterations are send to soft-demodulation module to get the likelihood of every bit for the decoder. The realization structure contains two main parts: Space-time Joint sub-module and Iteration Detection sub-module. The structure of Space-time Joint module here is similar with the DFT detector; Iteration Detection module contains 3 iteration sub-modules, which are parallel in space and serial in the processing of one sub-slot; Soft-demodulation is included in the last iteration sub-module.In the end of the thesis, the software-simulation results and hardware-realization results for both detectors are provided to make a compare.