基于分支定界和有效集的信号检测算法

大规模多输入多输出（multi-input multi-output，MIMO）系统中，随着天线数量的增多，现有传统信号检测算法在高阶调制时不能很好地平衡系统的检测性能和算法复杂度。为了解决以上问题，基于二次规划（quadratic programming，QP）检测器应用了有效集法和具有可变二分法的深度优先分支定界算法，提出了一种适用于大规模MIMO高阶调制系统的低复杂度检测算法，并提出了一种修剪策略和引入了近似因子，改善了系统性能，在复杂度和性能之间进行了更好地折中。复杂度分析表明，所提出算法复杂度比QP算法和二阶QP算法高，但比传统分支定界算法要低。仿真分析结果表明，在收发天线均为32的大规模MIMO场景下；在256QAM调制、误码率（bit error ratio，BER）为10^-4时，比传统分支定界算法提升了约3 dB的性能增益，验证了算法对高阶调制的适应性。

Signal detection algorithm based on branch bound and active set method

In the massive MIMO system, with the increase of the number of antennas, the existing traditional signal detection algorithms cannot well balance the detection performance and complexity in high-order modulation. To solve the above problems, based on the quadratic programming detector, this paper proposes a low complexity detection algorithm for massive MIMO high-order modulation systems by using the efficient set method and the depth-first branch and bound algorithm with variable dichotomy. In this paper, a pruning strategy and an approximation factor are proposed to improve the system performance and make a better tradeoff between complexity and performance. Complexity analysis shows the complexity of this algorithm is higher than that of the quadratic optimization algorithm and quadratic optimization algorithm but lower than the branch and bound algorithm. Simulation results show that in the massive MIMO scenario with 32 transmit and receive antennas, the performance gain of the algorithm is about 3 dB higher than that of the traditional branch and bound algorithm when the quadrature amplitude modulation order is eight and the bit error rate is 10^-4, which verifies the adaptability of the algorithm to high-order modulation.