Date of Award

5-2015

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Legacy Department

Electrical Engineering

Advisor

Kuang-Ching Wang

Committee Member

Harlan Russell

Committee Member

Haiying Shen

Committee Member

Hongxin Hu

Abstract

Integrated with the function of sensing, processing, and wireless communication, wireless sensors are attracting strong interest for a variety of monitoring and control applications. Wireless sensor networks (WSNs) have been deployed for industrial and remote monitoring purposes. As energy shortage is a worldwide problem, more attention has been placed on incorporating energy harvesting devices in WSNs. The main objective of this research is to systematically study the design principles and technical approaches to address three key challenges in designing reliable and sustainable WSNs; namely, communication reliability, operation with extremely low and dynamic power sources, and multi-tier network architecture. Mathematical throughput models, sustainable WSN communication strategies, and multi-tier network architecture are studied in this research to address these challenges, leading to protocols for reliable communication, energy-efficient operation, and network planning for specific application requirements. To account for realistic operating conditions, the study has implemented three distinct WSN testbeds: a WSN attached to the high-speed rotating spindle of a turning lathe, a WSN powered by a microbial fuel cell based energy harvesting system, and a WSN with a multi-tier network architecture. With each testbed, models and protocols are extracted, verified and analyzed. Extensive research has studied low power WSNs and energy harvesting capabilities. Despite these efforts, some important questions have not been well understood. This dissertation addresses the following three dimensions of the challenge. First, for reliable communication protocol design, mathematical throughput or energy efficiency estimation models are essential, yet have not been investigated accounting for specific application environment characteristics and requirements. Second, for WSNs with energy harvesting power sources, most current networking protocols do not work efficiently with the systems considered in this dissertation, such as those powered by extremely low and dynamic energy sources. Third, for multi-tier wireless network system design, routing protocols that are adaptive to real-world network conditions have not been studied. This dissertation focuses on these questions and explores experimentally derived mathematical models for designing protocols to meet specific application requirements. The main contributions of this research are 1) for industrial wireless sensor systems with fast-changing but repetitive mobile conditions, understand the performance and optimal choice of reliable wireless sensor data transmission methods, 2) for ultra-low energy harvesting wireless sensor devices, design an energy neutral communication protocol, and 3) for distributed rural wireless sensor systems, understand the efficiency of realistic routing in a multi-tier wireless network. Altogether, knowledge derived from study of the systems, models, and protocols in this work fuels the establishment of a useful framework for designing future WSNs.

Share

COinS