A Web-based evolutionary model for Internet data caching

Abstract

Caching is a standard solution to the problem of insufficient bandwidth caused by the rapid increase of information circulation across the Internet. Cache consistency mechanisms are a crucial component of each cache scheme influencing the cache usefulness and reliability. The paper presents a model for optimizing Internet cache content by the use of a genetic algorithm and examines the model by trace-driven experiments. Cached data are considered as a population evolving over simulated time by a number of successive cache "generations". The model is tested by the use of traces provided by a Squid proxy cache server. Using trace-driven caching, we show that the proposed evolutionary mechanisms improve cache nonstaleness and consistency and result in an updated cache content.