On-Board Content Caching With Dynamic Cache Reconfiguration in Multi-Layer Satellite Edge Networks

Satellite Edge Computing (SEC) integrated with multi-layer networks is a promising solution for on-board content caching to meet the stringent content delivery requirements in the future network. However, designing an efficient on-board content caching strategy is challenging in satellite edge networks because of limited on-board resources, dynamic topology, and fluctuating traffic demand. Frequent cache updates are required to adapt to these dynamic conditions, but incur cache reconfiguration overhead, which can overload feeder links and increase cache update delays. To address these challenges, in this paper, we propose an on-board dynamic cache reconfiguration strategy that maximizes the cache hit rate while minimizing reconfiguration overhead. We design a proximity-based content popularity model and an Age of Information (AoI)-aware caching strategy to optimize on-board satellite resources, enhancing the cache hit rate and ensuring content freshness. We employ a greedy algorithm for an initial and fast solution for caching contents and enhance it with a Genetic Algorithm (GA)-based approach. Simulations with realistic datasets demonstrate that the proposed solution outperforms the Cooperative Content Caching (CCC) benchmark by 11.63% and 9.25% in terms of cache hit rate and feeder link load, respectively.