Wireless content caching has recently been considered as an efficient way in fog radio access networks (F-RANs) to alleviate the heavy burden on capacity-limited fronthaul links and reduce delivery latency. In this paper, an advanced minimal delay association policy is proposed to minimize latency while guaranteeing spectral efficiency in F-RANs. By utilizing stochastic geometry and queueing theory, closed-form expressions of successful delivery probability, average ergodic rate, and average delivery latency are derived, where both the traditional association policy based on accessing the base station with maximal received power and the proposed minimal delay association policy are concerned. Impacts of key operating parameters on the aforementioned performance metrics are exploited. It is shown that the proposed association policy has a better delivery latency than the traditional association policy. Increasing the cache size of fog-computing based access points (F-APs) can more significantly reduce average delivery latency, compared with increasing the density of F-APs. Meanwhile, the latter comes at the expense of decreasing average ergodic rate. This implies the deployment of large cache size at F-APs rather than high density of F-APs can promote performance effectively in F-RANs.

中文翻译：

---

雾无线接入网络中遍历速率和传输延迟之间的权衡

无线内容缓存最近被认为是雾无线电接入网络 (F-RAN) 中的一种有效方式，可减轻容量受限的前传链路的沉重负担并减少交付延迟。在本文中，提出了一种先进的最小延迟关联策略，以最大限度地减少延迟，同时保证 F-RAN 中的频谱效率。利用随机几何和排队论，推导出成功投递概率、平均遍历速率和平均投递延迟的闭式表达式，其中基于以最大接收功率接入基站的传统关联策略和所提出的最小延迟关联政策有关。利用了关键操作参数对上述性能指标的影响。结果表明，所提出的关联策略比传统的关联策略具有更好的传递延迟。与增加 F-AP 的密度相比，增加基于雾计算的接入点 (F-AP) 的缓存大小可以更显着地降低平均交付延迟。同时，后者是以降低平均遍历率为代价的。这意味着在 F-AP 上部署大容量缓存而不是高密度的 F-AP 可以有效提高 F-RAN 的性能。