Bridging the gap between multiplexing and diversity in finite SNR multiple antenna coded caching

We study the multi-antenna Coded Caching setting in the finite Signal-to-Noise-Ratio (SNR) region, with the purpose to utilize the extra resources from the multiple antennas to increase the rate performance of the system. Specifically, assuming an L-antenna setting with K cache-enabled users (each with normalized cache-size γ), previous works showed that at the high SNR region the performance is proportional to L + Kγ degrees-of-freedom while, if one considers the low SNR region this would be outperformed by a strategy that focuses on beamforming a (single) multicast message (of Kγ + 1 elements) to its recipients. In order to bridge these two extremes we use a recently proposed algorithm that can balance the amount of users served in order to maximize the achieved sum-rate.