基于改进蚁群算法的大核矩阵搜索方案  

Large-kernel Matrix Search Scheme Based on Improved Ant Colony Algorithm

作　　者：胡凌峰 黄志亮[1] 张莜燕 周水红[1] HU Lingfeng;HUANG Zhiliang;ZHANG Youyan;ZHOU Shuihong(School of Physics and Electronic Information Engineering,Zhejiang Normal University,Jinhua Zhejiang 321004,China)

机构地区：[1]浙江师范大学物理与电子信息工程学院,浙江金华321004

出　　处：《通信技术》2025年第1期14-18,共5页Communications Technology

摘　　要：大核极化码相较于核矩阵维度更小的极化码,通常有着更大的极化速率,所以有着更优的译码性能。然而,随着维度的增加,核矩阵的搜索空间和极化速率的计算复杂度呈指数级增加,但现有的研究并不能摆脱维度增加所带来的指数级复杂度。引入智能优化算法——蚁群算法来搜索较大极化速率的大核矩阵,通过适当调整算法参数,算法能在可行时间内找到较优的大核矩阵。另外,引入莱维飞行进行算法优化,避免算法过早陷入局部最优。实验结果表明,所提算法能够稳定地找出13阶及以下拥有最佳极化速率的核矩阵,对于更高维度的矩阵也能输出不错的结果。Compared to polar codes with smaller kernel matrix dimensions,large-kernel polar codes usually have larger polarization rates and therefore have better decoding performance.However,with the increase of dimension,the search space of the kernel matrix and the computational complexity of the polarization rate increase exponentially,but the existing research cannot get rid of the exponential complexity brought by the increase of dimension.For the first time,an intelligent optimization algorithm,the ant colony algorithm,is introduced to search for large-kernel matrices with large polarization rate,and by appropriately adjusting the parameters of the algorithm,a superior large-kernel matrix could be found within a feasible time.In addition,Lévy flight is introduced to prevent the algorithm from falling into local optimum too early.Experimental results indicate that the proposed algorithm can stably find out kernel matrices with the best polarization rates of order 13 or less,and can also output good results for matrices of higher dimensions.

关 键 词：极化码 极化速率 蚁群算法 莱维飞行 最优核矩阵 

分 类 号：TN911.22[电子电信—通信与信息系统]