Current browse context:
cs
Change to browse by:
References & Citations
Computer Science > Information Theory
Title: Online Caching with Optimal Switching Regret
(Submitted on 18 Jan 2021)
Abstract: We consider the classical uncoded caching problem from an online learning point-of-view. A cache of limited storage capacity can hold $C$ files at a time from a large catalog. A user requests an arbitrary file from the catalog at each time slot. Before the file request from the user arrives, a caching policy populates the cache with any $C$ files of its choice. In the case of a cache-hit, the policy receives a unit reward and zero rewards otherwise. In addition to that, there is a cost associated with fetching files to the cache, which we refer to as the switching cost. The objective is to design a caching policy that incurs minimal regret while considering both the rewards due to cache-hits and the switching cost due to the file fetches. The main contribution of this paper is the switching regret analysis of a Follow the Perturbed Leader-based anytime caching policy, which is shown to have an order optimal switching regret. In this pursuit, we improve the best-known switching regret bound for this problem by a factor of $\Theta(\sqrt{C}).$ We conclude the paper by comparing the performance of different popular caching policies using a publicly available trace from a commercial CDN server.
Submission history
From: Samrat Mukhopadhyay [view email][v1] Mon, 18 Jan 2021 12:47:22 GMT (95kb,D)
Link back to: arXiv, form interface, contact.