A GA-based cache replacement policy

Caching has been recognized as an effective scheme for avoiding service bottleneck and reducing network traffic in World Wide Web. Cache replacement policy plays a key role in Web caching. This paper proposes a method based on genetic algorithm (GA) to provide a novel way for efficient cache replacement. A hybrid algorithm is chosen to work as fitness function. The cache replacement policy considers not only download time but also number of references to an object and object size. Experimental result shows that the GA-based cache replacement policy is effective.