Developing optimal usage plans for wireless sensor networks in smart grid

Wireless Sensor Networks (WSNs) are an enabling technology for both Smart Grid (SG) and Internet of Things (IoT) applications, like many other fields, due to their low cost and ease of installation, as well as the advanced communication capabilities they offer. Self-sustaining SGs, which allow two-way data communication between their components, are proposed as a remedy for the shortcomings of the conventional electrical grid. In this framework, it is critical to design energy efficient communication protocols to achieve a long network lifetime while maintaining the desired coverage level. In this thesis, we consider the scenario where the transmission paths are re-adapted to topology changes as long as the current set of sensors with positive residual energy ensures the required coverage level and connectivity to the base station. To this end, we propose a novel 0-1 mixed integer programming (MIP) model and two alternative 0-1 MIP models that optimize network behavior in unit time decomposition with an emphasis on energy consumption. The study is unique since topology adaptation due to energy depletion is managed within a holistic solution framework based on optimization methods. To this end, we avoid the traditional network lifetime metric, the time until the First Sensor Dies - FSD, and allow the WSN to operate optimally as long as a predetermined coverage level and connectivity to the BS can be maintahed with the remaining sensors. Accordingly, we observe that network lifetime can be increased significantly even after a one-time reorganization. Besided, we also analyzed how the solutions of the proposed models differ in terms of network lifetime and the ensured coverage level. As a side contribution, for the classical one-period problem, we emprically show that maximizing the time until the first sensor dies is equivalent to minimizing the energy consumption of the most energy consuming sensor, which is not necessarily true in the case of any topology change. Hence, the study provides a practical tool to determine the theoretical upper bound on the network lifetime with coverage and connectivity-based QoS requirements. SGs, smart transportation, smart homes, smart health, and smart monitoring/imaging applications are all based on the smarter use of sensors, which have been in our lives for many years. Thanks to sensors, the common vision of all these systems come together under the roof of IoT. Characteristic IoT quality of service requirements, based on criteria such as connectivity, coverage, and low latency, are mainly related to delay and network lifetime, and they can be tailored to the needs of the specific application area. Since they trade-off with each other, it is critical to balance them in order to develop quality designs in many respects. Therefore, it is meaningful to consider the effective WSN design problem from a multi-objective perspective considering coverage. In the second part of the thesis, we propose a 0-1 MIP based two-stage solution approach. For this purpose, we develop three different models, emphasizing maximum latency, average latency, and maximum energy consumption. Moreover, we examine the trade-off between network lifetime, data latency, and coverage. Therefore, we provide insight into the requirements of the different application areas. Keywords: WSN, Smart grid, IoT, Coverage, Network lifetime, Delay, Topology change, Multi-period optimization, Multi-objective optimization, Integer programming