Multiple-Symbol Differential Detection of M-ary Amplitude and Phase Shift Keying with Channel Estimation

(4，12)提出了16种APSK的环形信号集星座，并在高信噪比条件下最大化映射信号点之间的最小欧氏距离，给出了最佳环比。但为了在低信噪比下获得最佳的性能，需要利用其误码率(SER)来计算最佳的环比。为此，我们从理论上分析了理想相干检测、包络幅度/相干相位检测和包络幅度/差分相位检测三种检测方案的误码率，提出了一种基于最大似然准则的APSK多符号差分检测方法，并通过似然函数给出了它的判决度量。采用该检测方案的16个APSK的误码率性能明显好于以往文献中给出的近似度量。为了减少接收器…中的复杂性和计算量此外，在该检测方案中，将(M，L)-算法应用于树路径搜索。结果表明，在一个实用的接收机结构下，可以选择较大的待估计接收信号的序列长度，并且在最佳环比的情况下，其误码率性能与理想相干检测的性能几乎相同。最后，我们提出了一种在频率选择性衰落(FSF)信道中具有优异性能的自适应MSDD方案。自适应MSDD方案由一个自适应非相干最小均方信道估计器和MSDD组成，该估计器可以在没有任何判决延迟的情况下完成信道估计。由于自适应信道估计器和M算法的应用，这种自适应MSDD使得在没有任何判决延迟的情况下对M个存活路径上的每个符号进行信道估计成为可能。由于使用了MSDD和非相干信道估计器，因此这种自适应MSDD的误码率性能不受信道引起的相位波动的影响。将自适应MSDD方案应用于典型的16个APSK星座、(4，12)QAM星座和星型QAM星座。计算机仿真结果表明，该自适应MSDD方案在FSF信道下具有良好的跟踪性能。较少

(4,12) Circular-Signal-Set constellation has been proposed for 16 APSK and its optimum ring ratio has been given by maximizing minimum Euclidean distance between mapped signal points in high channel SNR. But in order to get the best performance in low SNR, the optimum ring ratio should be calculated by using its symbol error rate (SER). Thus, we analyze its symbol error rates theoretically for several detection schemes, which are the ideal coherent detection, the envelope amplitude/coherent phase detection, and the envelope amplitude/differential phase detection.Multiple-symbol differential detection (MSDD) based on the maximum likelihood (ML) criterion is proposed for APSK and its decision metric is given through the likelihood function, which is derived with no assumption. The SER performance of 16 APSK with this detection scheme shows better than that by using the approximated metric given in the past literature. To reduce the complexity and the computational efforts in the receiver … More , the (M, L)-algorithm is applied to tree path search in this detection scheme. As the result, the sequence length of received signals to be estimated can be chosen quite large under a practicable receiver structure, and we confirm that the SER performance shows almost the same as that of the ideal coherent detection with the optimum ring ratio.Lastly we propose a novel adaptive MSDD scheme that has excellent performance over frequency selective fading (FSF) channels. The adaptive MSDD scheme consists of an adaptive non-coherent least mean square channel estimator that can accomplish channel estimation without any decision delay and the MSDD. Because of the application of the adaptive channel estimator and the M-algorithm, this adaptive MSDD make it possible that channel estimation is accomplished for every symbol along M surviving paths without any decision delay. And the SER performance of this adaptive MSDD is not affected by phase fluctuation introduced by a channel because the MSDD and the non-coherent channel estimator are applied. The adaptive MSDD scheme is applied to typical constellation of 16 APSK, the (4,12) QAM and the star QAM. The excellent tracking performance of this adaptive MSDD scheme over FSF channels is confirmed by computer simulations. Less