通信学报2026，Vol.47Issue(2)：73-82,10.DOI:10.11959/j.issn.1000-436x.2026037
基于MMSE准则的鲁棒ICA-R算法初始化轻量化设计
Lightweight initialization design for robust ICA-R algorithm using MMSE criterion
摘要
Abstract
To enhance the blind signal separation performance of independent component analysis with reference(ICA-R)in communication spectrum-sharing applications and to reduce its hardware implementation complexity,a simplified non-random separation vector initialization method based on the minimum mean square error(MMSE)criterion was pro-posed,aiming at a lightweight design of a robust ICA-R algorithm.In this method,the observation signal matrix was di-rectly multiplied by the reference signal to compute the initial value of the separation vector.Theoretical analysis and hard-ware simulation results show that,compared with existing methods,the proposed approach offers comprehensive advan-tages including high separation accuracy,strong robustness,fast convergence,and low hardware overhead.Experimental results on the separation and extraction of spectrally overlapped communication signals using the ICA-R algorithm demon-strate that the proposed method can still achieve efficient and robust target signal separation and extraction under strong co-channel interference,while effectively reducing the algorithm runtime.Thus,it is suitable for resource-constrained commu-nication devices with stringent requirements for real-time performance,robustness,and computational efficiency.关键词
频谱共享/盲信号分离/参考独立分量分析/分离向量初始化Key words
spectrum sharing/blind signal separation/independent component analysis with reference/separation vector initialization分类
信息技术与安全科学引用本文复制引用
熊锦添,叶淦华,谢世珺,邓文,魏鹏,梁豪..基于MMSE准则的鲁棒ICA-R算法初始化轻量化设计[J].通信学报,2026,47(2):73-82,10.基金项目
国家自然科学基金资助项目(No.62201596) （No.62201596）
国防科技大学自主科研基金资助项目(No.ZK22-45) The National Natural Science Foundation of China(No.62201596),The Innovation Research Foundation of Na-tional University of Defense Technology(No.ZK22-45) （No.ZK22-45）
