ENERGY HARVESTING WIRELESS MULTIMEDIA SENSOR NETWORKS IN INDUSTRIAL ENVIRONMENTS

Abstract

Providing energy efficient and reliable communication for Industrial Wireless Sensor Networks (IWSNs) is of great significance when considering the harsh channel characteristics of industrial environment. However, prolonging a network lifetime while ensuring reliability becomes a major challenge. The main goal of this thesis is to maximize the network lifetime of Industrial Wireless Sensor Networks (IWSNs). The Energy Harvesting (EH) methods based on indoor solar, thermal and vibration that are suitable for industrial environments are defined and their contributions on network lifetime are investigated. A novel Mixed Integer Programming (MIP) model is formulated to maximize network lifetime by jointly considering path loss, application reliability and EH methods. Furthermore, communication in Wireless Multimedia Sensor Networks (WMSNs) causes the expense of extra energy consumption due to its huge data size. Therefore, reducing huge data size before transmission becomes important. To this end, the impact of the data size reduction methods such as compressive sensing and image compression while considering energy dissipation of both communication and computation on industrial network lifetime is evaluated. On the other hand, to solve the MIP model in a feasible time is hard especially when the large amount of sensor nodes deployed in the network. Heuristic based optimization methods are developed to overcome the timecomplexity of MIP problem.