The Research Of Signal Detectors In Vector Channels

The problem of signal detection in DS-CDMA, MIMO, and xDSL can all be attributed to the field of signal detection in vector channels. Although they seem to be different, their mathematical models are consistent. So the signal detection methods between DS-CDMA and MIMO systems can be similar.In this paper, firstly, a novel low-complexity (O(KlogK)) OD(Ordered-Detection) algorithm in DS-CDMA is presented. It has the features as follows:when the cross-correlation between distinct pairs of spreading sequence is constant, the algorithm is equivalent to Maximum Likelihood Detection; when the cross-correlations are not equal, the performance of the algorithm drops as the non-diagonal elements of the correlation coefficient matrix deviate increasingly from each other. If the deviation is small, OD still has a very good performance. Besides that, two expansion strategies are also given and make OD still have a good performance for some correlation coefficient matrices whose non-diagonal elements have large deviation between each other.Secondly, a further study on a greedy algorithm based CDMA detector whose computational complexity is O (K2logK) is done in this paper. It can be concluded that it should be better to use Decorrelator as the first stage of Greedy Detector for its good capability to fight against MAI and near-far effect. Besides, this paper also proposes a parallel processing strategy to decrease the detection delay.Finally, this paper presents a greedy algorithm based detector in MIMO system. The simulation shows that the algorithm has excellent performance in MIMO-OFDM system.