JISE

Fog Radio Access Network (F-RAN) has emerged as a promising architecture to reduce latency and network congestion by caching popular content at the edge. However, optimizing caching strategies in F-RAN faces challenges due to limited storage capacity and global vs. local content popularity. This paper proposes a novel proactive caching placement strategy using queueing theory to minimize latency. Specifically, we formulate an integer linear program based on a queuing model that captures content popularity and service rates. To solve this problem efficiently, we design two low-complexity heuristic algorithms: (1) An improved greedy algorithm that prioritizes globally popular content; and (2) A knapsack algorithm that optimizes cache allocation based on localized content popularity. Extensive simulations demonstrate that our proposed strategy achieves lower average latency and traffic cost compared to baseline caching schemes like LRU, LFU, and random replacement. The key innovation lies in optimizing caching decisions based on joint modeling of queuing delays and localized content popularity. This work provides an effective proactive caching framework for latency-critical F-RAN applications.