Topology Control in Wireless Sensor Networks: What Blocks the Breakthrough?

Abstract

Graph-based topology control adapts wireless topologies to achieve certain target graph structures. Wireless sensor networks seem well-suited for the expectations (in particular those on provided energy savings) raised by topology control. Nevertheless, topology control has never made the breakthrough in real-world deployments. This work explores the reasons for this, identifying five practical obstacles of today's topology control: (i) unrealistic assumptions, (ii) unsuitable graph structures, (iii) application agnosticism, (iv) unclear role in the stack, and (v) insufficient framework support. To address the latter obstacle, we provide a re-usable framework for the implementation and evaluation of topology control. Based on this framework, we conduct a testbed-based evaluation for two application scenarios and three topology control algorithms including a novel application-specific algorithm. Indeed, the identified obstacles hinder topology control from boosting the application. However, the achieved graph structures show the practical feasibility of topology control in principle.