基于Max-Log更新的马尔科夫链蒙特卡洛MIMO检测增强算法
发布时间：2018-11-07 13:22
【摘要】：针对传统的马尔科夫链蒙特卡洛(MCMC)算法,提出了一种基于Max-Log更新的MCMC-MIMO检测算法。该算法采用了基于Max-Log更新的采样,可以有效产生收敛于后验概率(APP)分布的比特样本列表集合,同时可避免计算传统MCMC算法中的每比特概率分布。但是该检测算法在高信噪比下,采样过程会陷入锁死到局部最优态。在此基础上,提出了3个增强技术:1)抖动处理,对给定置信区间内的更新进行抖动处理;2)条件下重新初始化,对处在潜在锁死态的采样序列进行重新初始化;3)修剪饱和处理,利用球形译码算法中的修剪饱和技术来处理MIMO检测输出的对数似然信息(LLR)。仿真结果显示,基于Max-Log更新的MCMC增强算法能有效地解决陷入锁死的问题,从而提高系统性能并降低系统的计算复杂度。在复杂度为MMSE-PIC检测算法的90%的基础上,性能提高了2 d B。
[Abstract]:Aiming at the traditional Markov chain Monte Carlo (MCMC) algorithm, a MCMC-MIMO detection algorithm based on Max-Log update is proposed. The algorithm adopts the sampling based on Max-Log update, which can effectively generate the sample list set converging to the posterior probabilistic (APP) distribution, while avoiding the computation of the per bit probability distribution in the traditional MCMC algorithm. However, under the high SNR, the sampling process will be locked to the local optimal state. On this basis, three enhancement techniques are proposed: 1) jitter processing, 2) reinitialization of the sample sequence in a potential locked state, 2) reinitialization of the update within a given confidence interval. 3) pruning saturation processing, using pruning saturation technique in spherical decoding algorithm to deal with logarithmic likelihood information (LLR). Of MIMO detection output. Simulation results show that the MCMC enhancement algorithm based on Max-Log update can effectively solve the problem of locking, thus improving the performance of the system and reducing the computational complexity of the system. The complexity of the algorithm is 90% of that of the MMSE-PIC detection algorithm, and the performance is improved by 2 dB.
【作者单位】： 电子科技大学通信抗干扰国家级重点实验室;
【基金】：国家自然科学基金(6150010678,61371104)
【分类号】：TN919.3
，
本文编号：2316497
[Abstract]:Aiming at the traditional Markov chain Monte Carlo (MCMC) algorithm, a MCMC-MIMO detection algorithm based on Max-Log update is proposed. The algorithm adopts the sampling based on Max-Log update, which can effectively generate the sample list set converging to the posterior probabilistic (APP) distribution, while avoiding the computation of the per bit probability distribution in the traditional MCMC algorithm. However, under the high SNR, the sampling process will be locked to the local optimal state. On this basis, three enhancement techniques are proposed: 1) jitter processing, 2) reinitialization of the sample sequence in a potential locked state, 2) reinitialization of the update within a given confidence interval. 3) pruning saturation processing, using pruning saturation technique in spherical decoding algorithm to deal with logarithmic likelihood information (LLR). Of MIMO detection output. Simulation results show that the MCMC enhancement algorithm based on Max-Log update can effectively solve the problem of locking, thus improving the performance of the system and reducing the computational complexity of the system. The complexity of the algorithm is 90% of that of the MMSE-PIC detection algorithm, and the performance is improved by 2 dB.
【作者单位】： 电子科技大学通信抗干扰国家级重点实验室;
【基金】：国家自然科学基金(6150010678,61371104)
【分类号】：TN919.3
，
本文编号：2316497
本文链接：https://www.wllwen.com/kejilunwen/xinxigongchenglunwen/2316497.html
