Focused on the performance improvement of receiver detection algorithm using signal space diversity, three aspects in the original algorithm are modified based on vector Gaussian approximation algorithm, including the calculation methods of probability expectation and variance with regarding to each symbol, the information dependence relationship in soft interference cancellation of the given symbol, and calculation of the probability density function. Therefore, the complexity of the algorithm is reduced and the system performance to resist delay is improved. According to simulation results and performance analysis by the new modified algorithm, under the condition of Rayleigh channels while lattice dimension N=16 and iteration number Iter=6, the BER curve is close to the ideal boundary. When bit error rate BER = 10-5, the performance gap between the Rayleigh fading channel and the Gaussian channel is only 0. 2 dB. In high delay requirement system, there is no degrading space for BER when the iteration number is four.%针对采用了信号空间分集的通信接收机检测算法的性能提升,以向量高斯近似检测算法为基础,通过改进原算法中每个符号的概率期望和方差的计算方法,给定符号的软干扰消除计算的信息依赖关系以及概率密度函教的计算,以降低算法的复杂度,提升系统的抗衰落能力.对改进算法进行了仿真比较和性能分析.结果表明,在瑞利衰落信道中,当旋转矩阵维数N=16,迭代次数为6时的误比特率曲线与理想边界接近,在误比特率BER=10-5时,仅相差0.2 dB；在时延要求较高的通信系统应用中,迭代4次即可满足要求,且继续增加迭代次数已没有误比特率下降裕量.
展开▼
