Thesis title: Three Branch Diversity Systems for Multi-Hop IoT Networks

Dissertation Presented by: Sakil Ahmed Chowdhury

The University of Vermont

Defense Date: May 4, 2022 at 7:30 pm

Thesis Examination Committee:

Jeff Frolik, Ph.D., University of Vermont, USA, Advisor

Abderrahim Benslimane, Ph.D., Avignon University, France, Co-advisor

Safwan Wshah, Ph.D., University of Vermont, USA

Tian Xia, Ph.D., University of Vermont, USA

Laurent Hébert-Dufresne, Ph.D., Chairperson, University of Vermont, USA

Cynthia Forehand, Ph.D., Dean of the Graduate College, University of Vermont, USA

Abstract:

Internet of Things (IoT) is an emerging technological paradigm connecting numerous smart objects for advanced applications ranging from home automation to industrial control to healthcare.

The rapid development of wireless technologies and miniature embedded devices has enabled IoT systems for such applications, which have been deployed in a variety of environments. One of the factors limiting the performance of IoT devices is the multipath fading caused by reflectors and attenuators present in the environment where these devices are deployed. Leveraging polarization diversity is a well-known technique to mitigate the deep signal fades and depolarization effects caused by multipath. However, neither experimental validation of the performance of polarization diversity antenna with more than two branches nor the potency of existing antenna selection techniques on such antennas in practical scenarios has received much attention. The objectives of this dissertation are threefold. First, to demonstrate the efficacy of a tripolar antenna, which is specifically designed for IoT devices, in harsh environments through simulations and experimental data. Second, to develop antenna selection strategies to utilize polarized signals received at the antenna, considering the restrictions imposed due to resource limitations of the IoT devices. Finally, to conduct comparative analyses on the existing standard diversity techniques and proposed approaches, in conjunction with experimental data.

Accordingly, this dissertation presents the testing results of tripolar antenna integrated with Arduino based IoT devices deployed in environments likely to be experienced by IoT devices in real life applications. Both simulation and experimental results from single point-to-point wireless links demonstrate the advantage of utilizing tripolar antennas in harsh propagation conditions over single branch antenna. Motivated by these empirical results, we deployed a small-scale IoT network with tripolar antenna based nodes to analyze the impact of tripolar antenna on neighbor nodes performance as well as to investigate end-to-end network performance. Our work illustrates that the selection of antenna branches while considering network architecture and the level of congestion on the repeater nodes minimizes excessive antenna switching and energy consumption.

Similar results are shown for IoT networks with predetermined and dynamic routing protocols, where the proposed techniques yielded lower energy consumption than the conventional diversity schemes. Furthermore, a probabilistic, low complexity antenna selection approach based on Hidden Markov model is proposed and implemented on wireless sensor nodes aiming to reduce energy consumption and improve diversity gain. Finally, the work presents a dual-hop based technique where a node selects the antenna element for optimal performance based on its immediate network neighbors antenna configuration status during selection. The performance of the proposed technique, which is verified through simulation and measured data, highlights the importance of considering network-wide evaluations of antenna selection techniques.

—