A Resource Allocation Perspective on Caching to Achieve Low Latency

The rapid growth of Internet contents from wireless data networks, especially social media, results in unprecedented traffic volume, invoking a challenge to the load of cellular infrastructure. Moreover, instead of traditional quality of service (QoS), the demands of quality of experience (QoE) become a more practical norm, which could barely be improved under present resource allocation on radio spectrum. With this in mind, we more generally consider resource allocation by exploiting caching in radio access networks (RANs), and show that caching, as a sort of storage, could be viewed as a substitution of the communication spectrum. Thus, we propose a collaborative strategy to implement caching in infrastructure and in mobile devices simultaneously, which in general turns out to be device-to-device (D2D) communication. This new design paradigm enables great reduction of latency for requesting Internet contents and can be implemented via slight amendments to present cellular systems.