Federated learning based QoS-aware caching decisions in fog-enabled internet of things networks

• DOI: 10.1016/j.dcan.2022.04.022
• Publication Date: 2022-04-30T12:02:48Z
• Authors: Xiaoge Huang; Zhi Chen; Qianbin Chen; Jie Zhang
• Keywords: IoT; D2D communication; Federated learning; 0202 electrical engineering, electronic engineering, information engineering; Fog computing network; Information technology; 02 engineering and technology; Deep neural network; T58.5-58.64

Quality of Service (QoS) in the 6G application scenario is an important issue with premise massive data transmission. Edge caching based on fog computing network considered as a potential solution to effectively reduce content fetch delay for latency-sensitive services Internet Things (IoT) devices. Considering time-varying scenario, machine learning techniques could further by optimizing decisions. In this paper, minimize and ensure QoS network, Device-to-Device (D2D) assisted architecture introduced, which supports federated QoS-aware decisions user preferences. To release congestion risk privacy leakage, learning, enabled D2D-assisted network. Specifically, it has been observed that yields suboptimal results according Non-Independent Identical Distribution (Non-IID) local users data. address issue, distributed cluster-based preference estimation algorithm proposed optimize placement, improve cache hit rate, convergence can mitigate impact Non-IID set clustering. The simulation show provides considerable performance improvement better compared existing algorithms.