Integrated Optimization of Heterogeneous-Network Management and the Elusive Role of Macrocells

Abstract

We consider heterogeneous wireless networks in the physical interference model and introduce a new formulation of the mixed-integer nonlinear programming (MINLP) problem that addresses base-station activation and many-to-many associations while minimizing power consumption. This formulation is complete, in the sense not only of tackling activations and associations in an integrated manner, but also in that the decoding possibilities allowed in the physical interference model are fully taken into account. The formulation’s integralities and resulting non-differentiabilities are all left in place as well, so solving the MINLP problem calls for somewhat unorthodox methods. Thus, we also introduce HetNetGA, a genetic algorithm that can tackle the problem as it comes, without any approximations. Though unsuitable for practical deployment, HetNetGA enables the investigation of such networks’ true possibilities, allowing for insights that more practical approaches cannot offer. In fact, the results we give, for scenarios involving both macrocells and picocells, though often aligned with what is expected, are sometimes unexpected and essentially point to the need to better understand the role of macrocells in helping provide capacity while remaining energetically advantageous. This contrasts sharply with previous approaches, which typically always spot situations in which macrocells can be accommodated. We expect HetNetGA, with its inherent ability to lead to results such as these, to be useful in helping identify and mitigate similar issues.

Keywords

• Capacity allocation
• heterogeneous wireless networks
• physical interference model
• power optimization