A Hybrid Approach Employing Coded Caching-Based Federated Learning (HA – CCFL) for Multiparameter Optimization in Cloud Frameworks

Abstract

Federated computing refers to the method of combining multiple distributed user groups and training models together without distributing data sets to client devices. In this chapter, the authors discussed a caching-based federated learning coded in fog computing called HS-CCFL, which is used as a content distribution artificial intelligence strategy with a single server connected to other devices through common constraints. They established an estimate of the long-term average speed of shared data for many users, each with a certain amount of memory. In this chapter, the authors show the experimental results by increasing the latency of fog points, which leads to an increase in latency and hit rate. Federation learning algorithm based on coded caching is discussed, where instead of local cache optimization, the closest term is added, which is related to the distance between local caching strategies. The experimental results justify the effectiveness of the method.