On exploiting network coding in cache-capable small-cell networks

Recently, network coding has emerged as an effective way to increase the efficiency of the content placement in the caching networks and thus boost the content delivery to the requesters. Its superiority compared to network coding-agnostic caching schemes lies on the increased availability of the content, which can be extracted by the requesters after they receive and decode a sufficiently large amount of encoded data. Although the topics surrounding network coding and caching have been already studied in the previous literature, their potential on enhancing mobile content delivery has not been fully explored yet. Namely, most of the existing works restrict the encoded data combinations to involve only parts of the same file, since this guarantees a low number of choices and thus simplifies the analysis. In this work, we study the problem of caching linear combinations of different files in a small-cell network. Our goal is to mitigate the pressure on the macrocellular base station by serving as many as possible content requests by the cache-endowed small-cell base stations that are deployed in the cell. Because of the NP-hardness of this problem, we propose a heuristic algorithm that gradually increases the performance of the obtained solution. Numerical results for typical popularity distributions reveal the performance benefits of our approach.