Data Distribution-Aware Online Client Selection Algorithm for Federated Learning in Heterogeneous Networks

Federated learning (FL) has received significant at-tention as a practical alternative to traditional cloud-centric ma-chine learning (ML). The performance (e.g., accuracy and conver-gence time) of FL is hampered by the selection of clients having non-independent and identically distributed (non-IID) data. In addition, a long convergence time is inevitable if clients with poor computation or communication capabilities participate in the FL procedure (i.e., the straggler problem). To minimize convergence time while guaranteeing high learning accuracy, we first formulate an optimization problem on client selection. As a practical solution, we devise a data distribution-aware online client selection (DOCS) algorithm. In DOCS, the FL server finds several clusters having near IID data and then uses a multi-armed bandit (MAB) technique to select the cluster with the lowest convergence time. The evalu-ation results demonstrate that DOCS can reduce the convergence time by up to 10% similar to 41% and improve the learning accuracy by up to 4% similar to 13% compared to the traditional client selection schemes.

Automated summary

Federated learning (FL) is an alternative to traditional cloud-centric machine learning (ML) that has gained attention. The performance of FL is affected by the selection of clients with non-independent and identically distributed (non-IID) data. To minimize convergence time and improve learning accuracy, an optimization problem is formulated and a data distribution-aware online client selection (DOCS) algorithm is proposed.