Degree Name

Doctor of Philosophy


School of Electrical, Computer and Telecommunications Engineering


Coverage is required by various applications that leverage Internet of things (IoT) networks to monitor one or more targets. A specific coverage requirement could be to ensure all targets, e.g., the ingress/egress of a building, are monitored by a sensor node for the longest time period possible. Another requirement could be coverage quality, which relates to the number of samples collected by sensor devices. The coverage quality or lifetime of targets, however, is limited by the energy of sensor devices. This is because the operational time of fixed or mobile sensor nodes, e.g., unmanned aerial vehicles (UAVs), is a function of their available energy. In this respect, many works have proposed to equip sensor nodes with energy harvesting (EH) capabilities.

Henceforth, this thesis aims to address targets coverage problems in EH IoT networks. First, it focuses on maximizing complete targets coverage lifetime using static devices. To this end, the problem at hand is to decide the set cover in each time slot, where each set cover ensures all targets are under the coverage of a device. In this respect, it presents a mixed integer linear program (MILP) to determine complete targets coverage. Further, it shows how the MILP can be represented as a factor graph, which can then be used by the belief propagation algorithm to compute set covers over time using the residual energy of devices.

The second aim focuses on an IoT network with UAVs and solar-powered charging stations. It considers K-coverage of targets, where at least K targets must be covered by a UAV. To this end, this thesis first formulates an UAVs assignment problem as an MILP. Although the MILP yields the optimal assignment, it does so using non-causal energy arrivals information at charging stations. To this end, this thesis proposes two solutions that use only causal energy arrivals information. In particular, these approaches employ model predictive control (MPC) or Monte Carlo tree search (MCTS). Further, they use a Gaussian mixture model (GMM) to estimate future energy arrivals.

Lastly, this thesis investigates the use of UAVs as aerial relays to boost coverage quality. The problem at hand is to decide the activation and data transmission rate of devices, and the location of UAVs. These quantities are optimized using an MILP. Further, this thesis proposes two heuristic solutions that decouple the MILP into different sub-problems; each sub-problem is then solved alternately. Moreover, the MPC solution relies only on causal energy arrivals information.

FoR codes (2020)

400608 Wireless communication systems and technologies (incl. microwave and millimetrewave)



Unless otherwise indicated, the views expressed in this thesis are those of the author and do not necessarily represent the views of the University of Wollongong.