Influence of Density on Throughput Performance in Cognitive Ultra-dense Networks

Abstract

Current advancements of Fifth Generation (5G) of mobile communications and beyond, have envisioned future networks as highly dense and coexisting in various bandwidths, providing seamless connectivity to users at any location. Thus, it is important to describe the effects and limits of densification and spectrum sharing. This article examines a less explored system model of a terrestrial cognitive radio (CR) based ultra-dense network (UDN) that operates within the range of a cellular macro base station (BS) and its users. It shares the incumbent spectrum in the interweave mode to avoid interference to the primary network, by implementing two common methods for energy detection (ED) spectrum sharing – Gaussian ED and Fading ED (FED). Through extensive simulations, the critical density of the UDN’s cognitive access points (CAPs), the ED efficiency, as well as the throughput gains, are determined through the measured signal-to-noise-ratio (SNR) at the CAPs and SUs. Additionally, the influence of different SU densification on the throughput is analyzed for the critical CAP density. It has been assessed that due to the high path loss in UDNs, the spectrum utilization gain (SUG) is small, but it may be improved through appropriate SU densification.