| To meet the requirement of future mobile communications, the work on wirelesscommunication systems strives for higher data rate all along. Although MIMO techniqueis widely recognized as one important approach to increase channel capacity, in high ratecondition, MIMO could be incapable of action in the aspects of energy efficiency anddetection complexity. To solve this problem, spatial modulation (SM), which is known asa single radio-frequency MIMO, is proposed. Compare with conventional MIMO systems,SM has high energy efficiency and low detection complexity, which make it a candidatetechnique for future wireless communication systems, especially for large-scale MIMO.This thesis bases on the5th generation (5G) wireless communication systems, aimsat better multi-antenna transmission, mainly studies SM receiver detector, some SMdetector enhance schemes are proposed, such as:1. Considering the layer search in MIMO signal detection, the tree search structuretailored to SM is proposed. The proposed structure regards each receive antenna as a levelof the tree so that the tree search becomes simple, there is no need to do QRdecomposition. This structure lays the foundation for application of tree search in SM.2. The complexity of SM with large-scale antennas or high-order signal constellationis still great, to solve this problem, the M-algorithm to maximum likelihood(M-ML)detector is proposed. This algorithm avoids exhaustive search by selecting only fixednumber of candidates. With the fixed complexity and breadth-first parallel structure,M-ML is conducive to practical implementation.3. Analyze the application of tree search in SM signal detection comprehensivelybases on depth-first and breadth-first. A first-layer sorting based Rx-SD algorithm isproposed to improve the SD method and a receive power based M-ML is proposed forM-ML. The simulation results show that the improved detectors can reduce thecomplexity further while maintaining a near-optimum performance. |
PDF Full Text Request