Skip to Main Content
Most of existing cache replacement policies are access-based replacement policies where update process is ignored. However, update information is extremely important. In this paper, we provide a deep analysis on cache access algorithms, and propose a step-wise optimal update-based replacement policy, called update-based step-wise optimal (USO) scheme, to optimize transmission cost and effective hit ratio at each replacement. Unlike traditional studies of replacement policies which are mostly based on only intuitions, our proposed scheme is based on quantitative analysis, and optimality is proved by an analytical model. The extensive simulations have shown that the advantage of the proposed replacement policy.