Coded caching with non-identical user demands

Abstract

Coded caching has gained a lot of interest recently. It jointly optimizes placement and delivery phases to create network-coded multicast opportunities to attain performance that surpasses conventional uncoded caching schemes. To the best of our knowledge, all coded caching systems studied in the literature are based on content popularity distributions as seen from the server's point of view. In this paper, we present a different approach to deal with coded caching problem. In particular, we analyze coded caching systems when each user's preferences are known to the system. Numerical results show that our approach outperforms the existing methods in certain regimes.