Tree pruning is an effective algorithm to reduce the complexity of sphere detection (SD) for multiple-input multiple-output (MIMO) communication systems. How to determine the tree pruning rule, as well as by what ...Tree pruning is an effective algorithm to reduce the complexity of sphere detection (SD) for multiple-input multiple-output (MIMO) communication systems. How to determine the tree pruning rule, as well as by what the tradeoff between the performance and the complexity can be achieved, is still an open problem. In this paper, a tree pruning algorithm is proposed based on minimum mean square error (MMSE) detection. The proposed algorithm first preforms MMSE detection since the complexity of MMSE detection is very low. Then the pruning constraints will be set according to the scaled path metrics of the MMSE solution. The choice of the scale factors and their influences on the complexity and performance are also discussed. Through analysis and simulations, it is shown that the complexity is reduced significantly with negligible performance degradation and additional computations.展开更多
基金supported by the Hi-Tech Research and Development Program of China (2011AA01A204)the Beijing University of Posts and Telecommunications Research and Innovation Fund for Youths
文摘Tree pruning is an effective algorithm to reduce the complexity of sphere detection (SD) for multiple-input multiple-output (MIMO) communication systems. How to determine the tree pruning rule, as well as by what the tradeoff between the performance and the complexity can be achieved, is still an open problem. In this paper, a tree pruning algorithm is proposed based on minimum mean square error (MMSE) detection. The proposed algorithm first preforms MMSE detection since the complexity of MMSE detection is very low. Then the pruning constraints will be set according to the scaled path metrics of the MMSE solution. The choice of the scale factors and their influences on the complexity and performance are also discussed. Through analysis and simulations, it is shown that the complexity is reduced significantly with negligible performance degradation and additional computations.