Achieving Spatial Scalability for Coded Caching via Coded Multipoint Multicasting

Abstract

The coded caching scheme proposed by Maddah-Ali and Niesen (MAN) critically hinges on the ability of the system to deliver a common coded multicast message from a server to all users in the system at a fixed rate, independent of the number of users. In order to apply this paradigm to a spatially distributed wireless network, it is important to make sure that such a common multicast rate does not vanish, as the number of users in the network and/or the network area increase. This paper starts from a variant of the MAN scheme successively proposed for the so-called combination network, where the multicast message is further encoded by a maximum distance separable (MDS) code, and the MDS-coded blocks are sent to multiple spatially distributed single-antenna edge nodes (ENs), transmitting at a fixed rate with no channel state information. The users have multiple antennas. They obtain receiver channel state information from the standard downlink pilots and can select to decode a desired number of EN transmissions while either nulling or treating as noise the others. The system is reminiscent of the so-called evolved Multimedia Broadcast Multicast Service, since the fundamental underlying transmission mechanism is multipoint multicasting, where each user can independently (in a user-centric manner) decide which EN to decode, without any explicit association of users with ENs. We study the performance of the proposed system when users and ENs are distributed according to homogeneous Poisson point processes in the plane, and the propagation is affected by Rayleigh fading and distance-dependent pathloss. Our analysis allows the optimization of the PHY parameters (PHY coding rate at the ENs and MDS coding rate) for given MAN scheme parameters. The proposed scheme achieves full spatial scalability in the following sense: for an extended network with arbitrary constant ratio of users per EN and area A = O(NE), where NE denotes the number of ENs, the system achieves a per-user delivery rate that does not vanish as N <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">E</sub> → ∞.