A Planning and Optimization Framework for Hybrid Ultra-Dense Network Topologies

Abstract

The deployment of small cells has been a critical upgrade in fourth-generation mobile networks as they provide macrocell traffic offloading gains, improved spectrum reuse and reduce coverage holes. The need for small cells will be even more critical in fifth-generation networks due to the introduction of higher spectrum bands, which necessitate denser network deployments to support larger traffic volumes per unit area. A network densification scenario envisioned for evolved fourth and fifth generation networks is the deployment of ultra-dense networks with small cell site densities exceeding 90 sites/km 2 (or inter-site distances of less than 112 m). The careful planning and optimization of ultra-dense networks topologies have been known to significantly improve the achievable performance compared to completely random (unplanned) ultra-dense network deployments by various third-party stakeholders (e.g. homeowners). However, these well-planned and optimized ultra-dense network deployments are difficult to realize in practice due to various constraints, such as limited or no access to preferred optimum small cell site locations in a given service area. The hybrid ultra-dense network topologies provide an interesting trade-off, whereby, an ultra-dense network may constitute a combination of operator optimized small cell deployments that are complemented by random small cell deployments by third-parties. In this study, an ultra-dense network multiobjective optimization framework and post-deployment power optimization approach are developed for realization and performance comparison of random, optimized and hybrid ultra-dense network topologies in a realistic urban case study area. The results of the case study demonstrate how simple transmit power optimization enable hybrid ultra-dense network topologies to achieve performance almost comparable to optimized topologies whilst also providing the convenience benefits of random small cell deployments.