Fundamental limits of caching: Improved rate-memory trade-off with coded prefetching

We consider a cache network, in which a single server is connected to multiple users via a shared error free link. The server has access to a database with N files of equal length F, and serves K users each with a cache memory of MF bits. A novel centralized coded caching scheme is proposed for scenarios with more users than files N ≤ K and cache capacities satisfying (1/K) ≤ M ≤ (N/K). The proposed scheme outperforms the best rate-memory region known in the literature if N ≤ K ≤ ((N2+ 1)/2).