Ascent Similarity Caching with Approximate Indexes

TL;DR

This paper introduces AÇAÍ, a similarity caching policy that leverages approximate indexes and a mirror ascent algorithm to efficiently serve large objects with low latency in edge computing environments.

Contribution

It proposes a novel similarity caching approach combining approximate indexing and a mirror ascent update method, enhancing performance over existing techniques.

Findings

Improved cache hit rates compared to previous methods.

Effective handling of non-stationary request patterns.

Strong theoretical guarantees for cache updates.

Abstract

Similarity search is a key operation in multimedia retrieval systems and recommender systems, and it will play an important role also for future machine learning and augmented reality applications. When these systems need to serve large objects with tight delay constraints, edge servers close to the end-user can operate as similarity caches to speed up the retrieval. In this paper we present A\c{C}AI, a new similarity caching policy which improves on the state of the art by using (i) an (approximate) index for the whole catalog to decide which objects to serve locally and which to retrieve from the remote server, and (ii) a mirror ascent algorithm to update the set of local objects with strong guarantees even when the request process does not exhibit any statistical regularity.