Adaptive Model Pruning and Personalization for Federated Learning over Wireless Networks
This addresses efficiency and accuracy challenges in federated learning for edge devices with non-IID data and limited wireless resources, representing an incremental improvement over existing personalization approaches.
The paper tackles the problems of decreased accuracy and increased latency in federated learning due to device heterogeneity and resource constraints by proposing a framework that splits models into pruned global and personalized parts, achieving approximately 50% reduction in computation and communication latency compared to baseline methods.
Federated learning (FL) enables distributed learning across edge devices while protecting data privacy. However, the learning accuracy decreases due to the heterogeneity of devices' data, and the computation and communication latency increase when updating large-scale learning models on devices with limited computational capability and wireless resources. We consider a FL framework with partial model pruning and personalization to overcome these challenges. This framework splits the learning model into a global part with model pruning shared with all devices to learn data representations and a personalized part to be fine-tuned for a specific device, which adapts the model size during FL to reduce both computation and communication latency and increases the learning accuracy for devices with non-independent and identically distributed data. The computation and communication latency and convergence of the proposed FL framework are mathematically analyzed. To maximize the convergence rate and guarantee learning accuracy, Karush Kuhn Tucker (KKT) conditions are deployed to jointly optimize the pruning ratio and bandwidth allocation. Finally, experimental results demonstrate that the proposed FL framework achieves a remarkable reduction of approximately 50 percent computation and communication latency compared with FL with partial model personalization.