Base Station Deployment and Interference Analysis for Urban Millimeter Wave Networks


Student thesis: Doctoral Thesis

View graph of relations


Related Research Unit(s)


Awarding Institution
  • Lin DAI (Supervisor)
  • Taejoon KIM (Supervisor)
Award date3 Mar 2020


Millimeter wave (mmWave) base stations (BSs) will be densely deployed to serve massive user equipments (UEs) in urban areas. Although interference from a single mmWave interferer is often low due to the severe propagation attenuation and directional transmission, the accumulated interference in dense mmWave networks can possibly be high. Simultaneously, network densification will incur significant BS installation and maintenance costs. Thus, in the presence of interference, a cost-efficient network deployment is essential.

In Chapter 2, this thesis provides an analytical stochastic geometry model that quantifies the total amount of mmWave network interference. Our study reveals that interference in mmWave networks is often larger than noise, but is generally lower than the desired power, yielding acceptable signal-to-interference-noise ratio (SINR) for link establishment.

In the context of network design which attempts to minimize the number of deployed BSs subject to the UE quality-of-service (QoS) constraints, the BS density optimization and site-specific deployment techniques have been popularly investigated in sub-6GHz systems. An underlying assumption in the sub-6GHz was the static and penetrable link propagation, which cannot be directly extended to the mmWave. In this thesis, we aim to devise BS deployment techniques in the mmWave bands.

In Chapter 3, we propose a minimum cost BS density optimization framework, based on the stochastic geometry. In this framework, we express the UE connectivity in a Manhattan-type urban geometry (MTUG) as a function of the BS density and transmit power. The minimum cost BS deployment problem is formulated by incorporating a tight lower bound of the analyzed UE connectivity expression. Solving the problem by a proposed algorithm resolves the best tradeoff between the BS density and transmit power.

In Chapter 4, we investigate the site-specific mmWave BS deployment problem, which finds the minimum number of sites from predetermined candidate locations for BS installation. UE placements in the considered geometry are characterized by collecting sufficient number of measured realizations rather than modeling with stochastic models. A scenario sampling approach is employed to quantitatively sample a small portion of the UE realizations to formulate a small-scale BS deployment problem, whose optimal solution is feasible to a specified majority of the overall measured UE realizations. The reduced-scale problem is optimally solved by a proposed low-complexity algorithm.

    Research areas

  • Millimeter wave, Interference analysis, Stochastic geometry, Base station deployment, Scenario sampling