To bridge the performance gap between original probability data association （PDA） algorithm and the optimum maximum a posterior （MAP） algorithm for multi-input multi-output （MIMO） detection, a grouped PDA （GP-...To bridge the performance gap between original probability data association （PDA） algorithm and the optimum maximum a posterior （MAP） algorithm for multi-input multi-output （MIMO） detection, a grouped PDA （GP-PDA） detection algorithm is proposed. The proposed GP-PDA method divides all the transmit antennas into groups, and then updates the symbol probabilities group by group using PDA computations. In each group, joint a posterior probability （APP） is computed to obtain the APP of a single symbol in this group, like the MAP algorithm. Such new algorithm combines the characters of MAP and PDA. MAP and original PDA algorithm can be regarded as a special case of the proposed GP-PDA. Simulations show that the proposed GP-PDA provides a performance and complexity trade, off between original PDA and MAP algorithm.展开更多
A method of MIMO channel tracking based on Kalman filter and MMSE-DFE is proposed. The Kalman filter tracks the time-varying channel by using the MMSE-DFE decision and the MMSE-DFE conducts the next decision by using ...A method of MIMO channel tracking based on Kalman filter and MMSE-DFE is proposed. The Kalman filter tracks the time-varying channel by using the MMSE-DFE decision and the MMSE-DFE conducts the next decision by using the channel estimates produced by the Kalman filter. Polynomial fitting is used to bridge the gap between the channel estimates produced by the Kalman filter and those needed for the DFE decision. Computer simulation demonstrates that this method can track the MIMO time-varying channel effectively.展开更多
不同基站和用户间由于距离变化等使传播时延不同而造成小区间的迟延干扰,严重恶化了系统性能。为此,探讨了如何利用SLNR(Signal to Leakage Noise Ratio)预编码抑制小区间的迟延干扰,给出了小区协作的系统模型及小区间迟延干扰分析。针...不同基站和用户间由于距离变化等使传播时延不同而造成小区间的迟延干扰,严重恶化了系统性能。为此,探讨了如何利用SLNR(Signal to Leakage Noise Ratio)预编码抑制小区间的迟延干扰,给出了小区协作的系统模型及小区间迟延干扰分析。针对存在迟延的LTE(Long Term Evolution)系统设计了一种基于SLNR准则的迟延SLNR预编码,以抑制小区间的迟延干扰,降低接收终端的复杂度。同时,对迟延SLNR预编码在LTE系统多基站协作环境下的性能进行了仿真。仿真结果表明,该方法对小区间的迟延干扰有较好的抑制作用,对不同天线配置性能均较理想,与接收端处理方案比较性能有所改进。虽然信道估计误差对性能有一定影响,但在可接受范围。展开更多
文章研究了多用户上行传输过程毫米波大规模多输入多输出(multi-input and multi-output,MIMO)系统的波束选择问题,提出了一种基于深度学习的波束选择方法。针对使用透镜的多用户毫米波大规模MIMO上行传输过程,提出一种面向波束选择的...文章研究了多用户上行传输过程毫米波大规模多输入多输出(multi-input and multi-output,MIMO)系统的波束选择问题,提出了一种基于深度学习的波束选择方法。针对使用透镜的多用户毫米波大规模MIMO上行传输过程,提出一种面向波束选择的深度学习框架,通过信道数据预先对神经网络进行离线训练,然后将实测信号输入训练好的神经网络在线预测信道直达径对应的波束,从而实现波束选择;基于该深度学习框架制定了具体的训练细则,采用柔性最大值交叉熵函数作为损失函数,使用自适应矩估计优化器优化神经网络参数。仿真结果表明,该文提出的基于深度学习的波束选择方法优于现有的正交匹配追踪方法。展开更多
基金Sponsored by the National Natural Science Foundation of China(60572120)
文摘To bridge the performance gap between original probability data association （PDA） algorithm and the optimum maximum a posterior （MAP） algorithm for multi-input multi-output （MIMO） detection, a grouped PDA （GP-PDA） detection algorithm is proposed. The proposed GP-PDA method divides all the transmit antennas into groups, and then updates the symbol probabilities group by group using PDA computations. In each group, joint a posterior probability （APP） is computed to obtain the APP of a single symbol in this group, like the MAP algorithm. Such new algorithm combines the characters of MAP and PDA. MAP and original PDA algorithm can be regarded as a special case of the proposed GP-PDA. Simulations show that the proposed GP-PDA provides a performance and complexity trade, off between original PDA and MAP algorithm.
文摘A method of MIMO channel tracking based on Kalman filter and MMSE-DFE is proposed. The Kalman filter tracks the time-varying channel by using the MMSE-DFE decision and the MMSE-DFE conducts the next decision by using the channel estimates produced by the Kalman filter. Polynomial fitting is used to bridge the gap between the channel estimates produced by the Kalman filter and those needed for the DFE decision. Computer simulation demonstrates that this method can track the MIMO time-varying channel effectively.
文摘不同基站和用户间由于距离变化等使传播时延不同而造成小区间的迟延干扰,严重恶化了系统性能。为此,探讨了如何利用SLNR(Signal to Leakage Noise Ratio)预编码抑制小区间的迟延干扰,给出了小区协作的系统模型及小区间迟延干扰分析。针对存在迟延的LTE(Long Term Evolution)系统设计了一种基于SLNR准则的迟延SLNR预编码,以抑制小区间的迟延干扰,降低接收终端的复杂度。同时,对迟延SLNR预编码在LTE系统多基站协作环境下的性能进行了仿真。仿真结果表明,该方法对小区间的迟延干扰有较好的抑制作用,对不同天线配置性能均较理想,与接收端处理方案比较性能有所改进。虽然信道估计误差对性能有一定影响,但在可接受范围。
文摘文章研究了多用户上行传输过程毫米波大规模多输入多输出(multi-input and multi-output,MIMO)系统的波束选择问题,提出了一种基于深度学习的波束选择方法。针对使用透镜的多用户毫米波大规模MIMO上行传输过程,提出一种面向波束选择的深度学习框架,通过信道数据预先对神经网络进行离线训练,然后将实测信号输入训练好的神经网络在线预测信道直达径对应的波束,从而实现波束选择;基于该深度学习框架制定了具体的训练细则,采用柔性最大值交叉熵函数作为损失函数,使用自适应矩估计优化器优化神经网络参数。仿真结果表明,该文提出的基于深度学习的波束选择方法优于现有的正交匹配追踪方法。
