摘要
In multi-user distributed antenna systems （DAS）, the scenario where each distributed antenna port is a multi-antenna array has not been comprehensively studied. In this article, four simple extended methods for downlink transmission are analyzed and compared in this scenario, two of which are based on the block diagonalization （BD） algorithm, namely joint BD and intra BD method. The other two methods are joint time division multiplexing （TDM） method and central antenna system （CAS） method. These four methods are evaluated through both theoretic analysis and insightful simulations. The theoretic analysis includes the receiving signal combination, post-processing signal to interference plus noise ratio （SINR） calculation. It shows that the intra BD method requires less channel state information at the transmitter （CSIT） and has lower computational complexity and process latency. Simulation results show that the joint BD method provides the best performance, and the intra BD method suffers little performance loss. Overall, the intra BD method is proved to be an optimal tradeoff which achieves high capacity, especially for the cell edge users, with relatively low complexity when power constraints are considered.
In multi-user distributed antenna systems （DAS）, the scenario where each distributed antenna port is a multi-antenna array has not been comprehensively studied. In this article, four simple extended methods for downlink transmission are analyzed and compared in this scenario, two of which are based on the block diagonalization （BD） algorithm, namely joint BD and intra BD method. The other two methods are joint time division multiplexing （TDM） method and central antenna system （CAS） method. These four methods are evaluated through both theoretic analysis and insightful simulations. The theoretic analysis includes the receiving signal combination, post-processing signal to interference plus noise ratio （SINR） calculation. It shows that the intra BD method requires less channel state information at the transmitter （CSIT） and has lower computational complexity and process latency. Simulation results show that the joint BD method provides the best performance, and the intra BD method suffers little performance loss. Overall, the intra BD method is proved to be an optimal tradeoff which achieves high capacity, especially for the cell edge users, with relatively low complexity when power constraints are considered.
基金
supported by Qualcomm and Projects (NSF60772112,2009ZX03003-008-01)
