Stack based detection algorithms for V-BLAST wireless system

A series of novel stack-based low-complexity algorithms are proposed for signal detection in V-BLAST wireless systems. Among them, the basic stack (BST) algorithm achieves maximum-likelihood (ML) detection performance with tractable computational complexity. By optimizing the BST algorithm through optimally ordering and adapting detection to channel matrix conditions, computational complexity and storage complexity are further reduced. An upper bound for the average computational complexity of the BST algorithm is derived analytically and it is shown that, for a fixed channel matrix, the upper-bound of the average computational complexity of the ML stack algorithm decreases exponentially as SNR increases. Simulation results show that stack-based algorithms achieve optimum performance with much lower average complexity than traditional ML detectors.