Joint Optimization of Resource Allocation and Data Selection for Fast and Cost-Efficient Federated Edge Learning

• DOI: 10.48550/arxiv.2407.02888
• Publication Date: 2024-07-03
• Authors: Yunjian Jia; Zhen Huang; Jiping Yan; Yulu Zhang; Kun Luo; Wanli Wen
• Keywords: FOS: Computer and information sciences; Computer Science - Machine Learning; Artificial Intelligence (cs.AI); Computer Science - Artificial Intelligence; Machine Learning (cs.LG)

Deploying federated learning at the wireless edge introduces (FEEL). Given FEEL's limited communication resources and potential mislabeled data on devices, improper resource allocation or selection can hurt convergence speed increase training costs. Thus, to realize an efficient FEEL system, this paper emphasizes jointly optimizing selection. Specifically, in work, through rigorously modeling process deriving upper bound one-round rate, we establish a problem of joint selection, which, unfortunately, cannot be solved directly. Toward end, equivalently transform original into solvable form via variable substitution then break it two subproblems, that is, problem. The subproblems are mixed-integer non-convex integer problems, respectively, achieving their optimal solutions is challenging task. Based matching theory applying convex-concave procedure gradient projection methods, devise low-complexity suboptimal algorithm for respectively. Finally, superiority our proposed scheme validated by numerical results.