A Survey of Edge Computing Resource Allocation Strategies Based on Federated Learning

Abstract

In the edge computing environment, the data on edge computing is at risk of leakage due to the variety and wide distribution of nodes. As a new distributed machine learning framework, federated learning can effectively solve the privacy and security problems of users' information data in different fields. However, in federated learning, with the continuous landing of AI applications and the growing demand for model reasoning services, the resources consumed by federated learning will exceed the computing power of edge computing, so it is necessary to study the resource allocation strategy of edge computing for federated learning. This paper first introduces the concepts of federated learning and edge computing, and resource allocation strategies based on edge computing; Then it introduces the challenges faced by federated learning and the operating system framework of edge computing based on federated learning; Secondly, it combs the resource allocation strategy of edge computing based on federated learning; Finally, the paper summarizes the work of the full text and analyzes the future development trend of resource allocation under Federated learning.