應用於多天線偵測之可變可重疊之叢集演算法

Abstract

在這篇論文裡，我們推薦一個具有高輸出率、固定複雜度的硬性輸出球體解碼器，並支援高維度64QAM、256QAM調變的4T4R和8T8R多輸入多輸出通訊系統。 本論文提出一基於可變可重疊之叢集的MIMO偵測方法(Variable and Overlapped Cluster-based MIMO Detection Algorithm)，在PER 在0.08下與最大相似算法誤差在0.5dB以內，較K best球體解碼器演算法低複雜度，可硬體實作的演算法。 本演算法是在偵測前，以基礎等化器找出的可能落點為依據，決策出該每個天線維度上可能候選星群，並以寬度優先搜尋的分支分界的方法，結合MMSE-SQRD解碼的系統架構來解碼，而本叢集分群法為以下兩種： 重疊叢集法(Overlap cluster)，為一減少複雜度並維持偵測效能之方法，在叢集的決策上，為了減少在邊界情況(Boundary condition)下的決策失誤，兩個不同的群可能會有相同的候選星群，藉此在叢集上有更高的準確性。 可變叢集法(Dynamic cluster)，為適應不同天線的通道衰減效應，而改進的叢集法。在決策的候選星群時，利用排序QR分解(Sorted QR Decomposition)演算法之通道路徑的範數(norm)資訊，使低通道衰減之天線有低數量之候選星群，高通道衰減天線維度擁有較多數量之候選星群候選星群，而此法符合實際天線陣列在真實情況下的傳輸環境。 實作於IEEE 802.11n的通訊平台上，提供4T4R和8T8R在高維度64QAM、256QAM調變，在符合TGN-E所規範的通道模型中進行模擬。模擬結果指出此演算法與傳統K-best球體解碼器，若維持PER 在0.08的誤差0.5dB之內，以較低的複雜度完成相同的系統效能；若維持約略相同複雜度之下，具有較佳的系統效能表現。因此，此演算法為多輸入多輸出系統提供了具有低複雜度、接近效能最佳化的偵測演算法。

Recently, multiple-input multiple-output (MIMO) architecture has been applied widely in many wireless communication systems because of its high spectrum efficiency. Various approaches are explored for the MIMO detection, the ZFD, the MMSED, V-BLAST, the maximum likelihood detection (MLD) as well as the Sphere Decode detection (SD). We propose the Variable and Overlapped Cluster-based MIMO Detection algorithm by partitioning the transmitted MIMO signal vectors into vary clusters with estimated symbol in each dimension in 64-QAM/256-QAM and finding out the result signal by comparing the received signal with all the candidates above. And the proposed method, step A) as well as B), are demonstrating in the following. In A), we demonstrate overlap clustering algorithm that the estimated signal got by linear detectors, as ZFD or MMSED, and then pick out the possible constellation points falling on each antenna according to the range which the estimated signal is in. After overlap clustering algorithm in step A, we enlarge/narrow the possible constellations points according to the column norm of H included channel gain information. Moving on B), we have all the candidates signals compare with the received signal, and then apply BFS with best K candidates in the searching space of MMSE SQRD. Eventually, the detection signal with the least accumulative square Euclidean distance is delivered. Through simulation in IEEE 802.11n platform with TGN channel E, it indicates the complexity of proposed algorithm is less than the K-best SD with the same performance. Hence, the proposed algorithm provides a near-optimal solution with low computation complexity design for wireless MIMO system.