Optimal 3D Cell Planning: A Random Matrix Approach

Axel Müller (1, 2) , Jakob Hoydis (3) , Romain Couillet (4) , Mérouane Debbah (1)

1 Chaire Radio Flexible - Chaire Radio Flexible Alcatel-Lucent/Supélec
2 IMC - Intel Mobile Communications
3 Bell Labs
4 E3S - Supélec Sciences des Systèmes

This article proposes a large system approximation of the ergodic sum-rate (SR) for cellular multi-user multiple-input multiple-output uplink systems. The considered system has various degrees of freedom, such as clusters of base stations (BSs) performing cooperative multi-point processing, randomly distributed user terminals (UTs), and supports arbitrarily configurable antenna gain patterns at the BSs. The approximation is provably tight in the limiting case of a large number of single antenna UTs and antennas at the BSs. Simulation results suggest that the asymptotic analysis is accurate for small system dimensions. Our deterministic SR approximation result is applied to numerically study and optimize the effects of antenna tilting in an exemplary sectorized 3D small cell network topology. Significant SR gains are observed with optimal tilt angles and we provide new insights on the optimal parameterization of cellular networks, along with a discussion of several non-trivial effects.

Domains: Computer Science [cs]; Networking and Internet Architecture [cs.NI]

Axel Müller, Jakob Hoydis, Romain Couillet, Mérouane Debbah. Optimal 3D Cell Planning: A Random Matrix Approach. 2012 IEEE Global Communications Conference (GLOBECOM), Dec 2012, Anaheim, California, United States. ⟨10.1109/GLOCOM.2012.6503829⟩. ⟨hal-00771220⟩