Fundamentals of Cluster-Centric Content Placement in Cache-Enabled Device-to-Device Networks

Abstract

This paper develops a comprehensive analytical framework with foundations in\nstochastic geometry to characterize the performance of cluster-centric content\nplacement in a cache-enabled device-to-device (D2D) network. Different from\ndevice-centric content placement, cluster-centric placement focuses on placing\ncontent in each cluster such that the collective performance of all the devices\nin each cluster is optimized. Modeling the locations of the devices by a\nPoisson cluster process, we define and analyze the performance for three\ngeneral cases: (i)$k$-Tx case: receiver of interest is chosen uniformly at\nrandom in a cluster and its content of interest is available at the $k^{th}$\nclosest device to the cluster center, (ii) $\\ell$-Rx case: receiver of interest\nis the $\\ell^{th}$ closest device to the cluster center and its content of\ninterest is available at a device chosen uniformly at random from the same\ncluster, and (iii) baseline case: the receiver of interest is chosen uniformly\nat random in a cluster and its content of interest is available at a device\nchosen independently and uniformly at random from the same cluster. Easy-to-use\nexpressions for the key performance metrics, such as coverage probability and\narea spectral efficiency (ASE) of the whole network, are derived for all three\ncases. Our analysis concretely demonstrates significant improvement in the\nnetwork performance when the device on which content is cached or device\nrequesting content from cache is biased to lie closer to the cluster center\ncompared to baseline case. Based on this insight, we develop and analyze a new\ngenerative model for cluster-centric D2D networks that allows to study the\neffect of intra-cluster interfering devices that are more likely to lie closer\nto the cluster center.\n