Blind sequence detection of QAM system with short received data based on Echo State Networks
Liu Shuan, Hengheng Li, Xiejie Jin, Xiukai Ruan, Tan Yanhua, Guijun Zhu, Qibo Cai, Zhili Zhou · 2016
A new blind sequence detection (BSD) approach using Echo State Network (ESN) is presented, together with the detailed theoretical derivation of the algorithm. The proposed method can be applied to those cases that the constellation of source signal is dense (e.g., 64QAM) and guarantees a convergence within a short data packet. Firstly, a ESN, in which the reservoir weight matrix is devised via QR decomposition from the view of receiving signal subspace which is different from the mechanism of generating those existed ESN weight matrices randomly, is constructed to solve the special issues of BSD. Simulation results verify the correctness and effectiveness of the new approach, and some special simulation phenomena of this algorithm are shown.