Device-to-Device Communication for Mobile Multimedia in Emerging 5G Networks
Jiajia Liu,Nei Kato,Hirotaka Ujikawa,Kenichi Suzuki
DOI: https://doi.org/10.1145/2983641
IF: 4.094
2016-01-01
ACM Transactions on Multimedia Computing Communications and Applications
Abstract:Device-to-device (D2D) communication, which utilizes mobile devices located within close proximity for direct connection and data exchange, holds great promise for improving energy and spectrum efficiency of mobile multimedia in 5G networks. It has been observed that most available D2D-based works—considered only the single-cell scenario with a single BS. Such scenario-based schemes, although tractable and able to illustrate the relationship between D2D links and cellular links, failed to take into account the distribution of surrounding base stations and user equipments (UEs), as well as the accumulated interference from ongoing transmissions in other cells. Furthermore, the single-tier network with one BS considered in available works is far from the real 5G scenario in which multi-tier BSs are heterogeneously distributed among the whole network area. In light of such observations, we present in this article a model for network coverage probability and average rate analysis in a D2D communication overlaying a two-tier downlink cellular network, where nineteen macro base stations (MBSs) with pico base stations (PBSs) placed at the end point of macro cell (hexagons) borders are employed according to the 3GPP specifications, and mobile users are spatially distributed according to the homogeneous Poisson Point Process model. Each mobile UE is able to establish a D2D link with adjacent UEs or connect to a nearby macro or pico base station. Stochastic geometric analysis is adopted to characterize the intratier interference distribution within the MBS-tier, PBS-tier, and D2D-tier based on which network coverage probability and per-user average rate are derived with a careful consideration of important issues such as threshold value, SINR value, user density, content hit rate, spectrum allocation, and cell coverage range. Our results show that, even for the overlaying case, D2D communication can significantly improve network coverage probability and per-user average downlink rate. Another finding is that the frequency allocation for D2D communications should be carefully tuned according to network settings, which may result in totally different varying behaviors for the per-user average rate.