Performance of positioning and lighting services for smart buildings : a wireless sensor network approach
Chan, Edwin Yiu Wing
Date of Issue2017-04-07
School of Electrical and Electronic Engineering
Continuous population growth and global urbanization is forcing us to review the current usage of resources, such as energy and water, in cities. In response, researchers and practitioners are leveraging the latest information and communication technologies (ICT) to make our cities smarter. Smart buildings in particular, as a building block of smart cities, will play a critical role in utilizing the resources in a more efficient and sustainable manner. A smart building, enabled by technologies, is able to monitor and control its subsystems effectively and efficiently, while at the same time, provide a comfortable and productive environment to its occupants. With the extensive use of wireless sensors networks (WSN) and Internet of Things (IoT) technologies within a building, real-time environment data are collected by sensors and relayed to the building automation system (BAS), which then holistically manages the services, e.g. lighting, HVAC and security, of the building. In this thesis, we focus on two important services, namely lighting and positioning, which are commonly found in smart buildings and can potentially benefit from the use of WSN technology. For the former, a WSN-based light management system, which involves sensing, control and wireless communications, is first studied. The issues of energy efficiency and occupant comfort are considered when a new model of daylight-responsive lighting and shading control problem is formulated. As a consequence of the preliminary evaluation, a survey of available illuminance sensing technologies is conducted. Subsequently, a wireless sensing platform consisting of high accuracy analog illuminance sensors is developed, and its performance is assessed in a high-tech life-sized test facility. Positioning service is another important asset to a smart building. The occupancy data, including positions and identities, that it obtains are critical to the effectiveness of other services, such as lighting and HVAC, in a smart building. Taking advantage of the envisaged large scale deployment of WSNs in smart buildings, the Discrete Weighted Centroid Localization algorithm (dWCL) is proposed as a potential solution for a WSN-based indoor positioning system (IPS). Under this architecture, a wireless device can estimate its own position within the building with the aid of the underlying WSN infrastructure. Considering the current state of digital electronics, dWCL is a more practical alternative to other WCL-based algorithms and is flexible to accommodate different performance requirements. Based on the presented analytical framework, two optimizing approaches are introduced to further elevate its performance. Finally, the thesis examines the fundamental limit of the class of range-free localization algorithms in WSN. Mathematical framework to analyze the resolution of this class of algorithms is proposed. In addition, the analytical expression for the localization error distribution is presented. Unlike the prior work based on the circular coverage model, a more realistic scenario is considered, where the radio propagation is subject to the log-normal shadow fading and the correlation of shadow fading is also taken into account in the analysis of the problem.
DRNTU::Engineering::Electrical and electronic engineering