通信学报2026，Vol.47Issue(4)：97-112,16.DOI:10.11959/j.issn.1000-436x.2026067
面向协作频谱感知的个性化差分隐私联邦学习方法
Personalized differential privacy federated learning method for collaborative spectrum sensing
摘要
Abstract
To address the degradation of model performance caused by data non-independent and identically distributed(Non-IID)characteristics in collaborative spectrum sensing,a federated learning scheme RebalFL was proposed,which integrated personalized differential privacy with a rebalancing clustering strategy.First,a personalized differential pri-vacy mechanism that allowed heterogeneous privacy budgets for different data sources was introduced,thereby reducing noise injection while preserving privacy.Then,a rebalancing clustering strategy was designed to form client clusters with more balanced data distributions and mitigate model drift.Experimental results show that RebalFL outperforms existing differential privacy methods in Non-IID scenarios,substantially improving the classification accuracy and robustness of spectrum sensing models under privacy protection.关键词
联邦学习/个性化差分隐私/频谱感知/隐私保护Key words
federated learning/personalized differential privacy/spectrum sensing/privacy protection分类
信息技术与安全科学引用本文复制引用
唐湘云,康嘉文,韩旭,张焘,刘寅秋,孙庚,焦雨涛..面向协作频谱感知的个性化差分隐私联邦学习方法[J].通信学报,2026,47(4):97-112,16.基金项目
国家自然科学基金资助项目(No.62572132,No.62572502,No.62571548) （No.62572132,No.62572502,No.62571548）
国家密码基金资助项目(No.2025NCSF02030) （No.2025NCSF02030）
国家自然科学基金青年科学基金资助项目(C类)(No.62302539,No.62402029) （C类）
北京市自然科学基金丰台联合基金资助项目(No.L251041) The National Natural Science Foundation of China(No.62572132,No.62572502,No.62571548),The National Cryptography Foundation of China(No.2025NCSF02030),NSFC Youth Science Fund Project(Category C)(No.62302539,No.62402029),Beijing Natural Science Foundation Fengtai Joint Fund(No.L251041) （No.L251041）
