Optimum Energy Efficiency and Area Spectral Efficiency Tradeoff in User-Centric Ultra-Dense Networks

The objective of this paper is to theoretically identify the optimum tradeoff between energy efficiency (EE) and area spectral efficiency (ASE) of a user-centric ultra-dense network (UC-UDN). In a UC-UDN, there are more remote access points (RAPs) than user equipments (UEs). Each UE is served by its closest RAP, and the RAPs that are not associated with any UE are put in sleeping mode. The RAPs and UEs are assumed to form independent homogeneous Poisson point process (PPP). Increasing RAP density can in general improve ASE due to shorter transmission distance, but usually at the cost of EE due to stronger mutual interference and higher energy consumption. Thus it is critical to identify the optimum RAP density that can achieve the optimum EE-ASE tradeoff. We tackle this problem by analytically solving an optimization problem that aims at maximizing the EE under the constraint of minimum achievable ASE. The solution to the problem provides a closed-form expression of the RAP density that can achieve the optimum EE-ASE tradeoff.