A multi-objective performance modelling framework for enabling self-optimisation of cellular network topology and configurations

Cellular system optimisation, a cornerstone of cellular systems paradigm, requires new focus shift because of the emergence of plethora of new features shaping the cellular landscape. These features include self-organising networks with added flavours of heterogeneity of cell sizes and base station types, adaptive antenna radiation patterns, energy efficiency, spatial homogeneity of service levels and focus shift from coverage to capacity. Moreover, to effectively tackle spatiotemporal dynamics of network conditions, a generic low-complexity framework to quantify the key facets of performance, that is, capacity, quality of service and energy efficiency of the various network topology configurations (NTC), is needed for enabling self-organising networks empowered cellular system optimisation on the fly. In this paper, we address this problem and present a performance characterisation framework that quantifies the multiple performance aspects of a given heterogeneous NTC through a unified set of metrics that are derived as function of key optimisation parameters and also present a cross comparison of a wide range of potential NTCs. Moreover, we propose a low-complexity heuristic approach for holistic optimisation of future heterogeneous cellular systems for joint optimality in the multiple desired performance indicators. The performance characterisation framework also provides quantitative insights into the new tradeoffs involved in optimisation of emerging heterogeneous networks and can pave the way for much needed further research in this area.