Adaptive sequence detection and information capacity improvement for wireless communication systems

Abstract

Increasing demand for wireless communication challenges the availability of limited radio resources, such as bandwidth and power etc. Limited resources cause a trade off with the quality of service. The work presented in this thesis is intended to develop algorithms those can be used to demodulate information with optimal amount of resources (Signal to Noise Ratio, Processing memory requirement, Computational complexity etc ). In the first part of the thesis a decision feedback sequence detection algorithm has been proposed that provides exactly the same bit error rate as in standard maximum likelihood sequence estimation but with 95% lowmomputational complexity. Besides that the proposed algorithm achieves 2 dB signal to noise ratio (SNR) gain over the existing decision feedback algorithms. The proposed algorithm is applicable in multiple input multiple output (MIMO) as well as single input single output (SISO) wireless communication systems. In the second part of the thesis an adaptive blind sequence detection algorithm has been proposed where a novel reference channel has been exploited. The problem of bit-shift ambiguity in blind sequence detection is completely eliminated exploiting the proposed algorithm. A 3 dB SNR gain is achieved against the existing blind sequence detection algorithms for the system without error correction code. The BER performance is highly scalable with the variation of segmentation window size. In the third part of the thesis, two different sequence detection algorithms have been proposed to track rapidly time varying channels. One of the algorithms, called extended window survivor processing (EWSP), requires lower computational complexity than that of Per-survivor processing (PSP) sequence detection process. The other algorithm, called bi-directional survivor processing reduces 17% of channel misacquisation than that of PSP. Consequently, both of these algorithms reduce the probability of error propagation in the detection process. In the final part of the thesis, the capacity and coverage of the UMTS urban network has been analyzed while the Repeaters are inserted. It has been found that the system capacity with repeaters is doubled in an environment with the propagation constant 3.7-3.8. As a by product, 10% increase in the cell coverage was also found.