Title:Uncoded Placement Optimization for Coded Delivery

Abstract:Existing coded caching schemes fail to simultaneously achieve efficient content placement for non-uniform file popularity and efficient content delivery in the presence of common requests, and hence may not achieve desirable average load under a non-uniform, possibly very skewed, popularity distribution. In this paper, we consider a network consisting of a server with access to N files and connected through a shared link to K users, each equipped with a cache. First, we present a class of coded caching schemes specified by a general file partition parameter reflecting the fraction of memory allocated to each file. Then, we formulate the coded caching optimization problem within the class with N2^K variables to minimize the average load under an arbitrary file popularity by optimizing the file partition parameter. Next, we impose some conditions on the file partition parameter, and derive a simplified average load expression. Based on the simplified expression, we transform the original optimization problem into a linear optimization problem with N(K + 1) variables under an arbitrary file popularity and a linear optimization problem with K + 1 variables under the uniform file popularity. Under the uniform file popularity, we also obtain a closed-form optimal solution. Finally, numerical results reveal that the imposed conditions do not affect the optimality of the original problem and demonstrate the performance gain of the optimized coded caching scheme over some existing schemes.