On the Modelling, Design, and Implementation of Distributed QoS-Aware Communication Protocols for Low-Power Internet of Things Networks

Abstract

Low-power Lossy Networks (LLNs) are important building blocks of the Internet of Things (IoT), where low-power resource-constrained devices use IPv6 end-to-end communication through short-range lossy wireless links. By connecting LLNs to the Internet, not only can low power devices be easily accessed and interacted with users or systems, but can also offload their processes and storage to remote cloud servers. This feature moved LLN to a new stage that includes a wide variety of applications such as home and building automation, industrial control, and smart cities. Due to the resource limitation of low-power IoT devices, conventional Internet protocols such as TCP are not suitable for implementation on LLNs. Since 2012, the Internet Engineering Task Force (IETF) organization has been designing standards for connecting low-power IoT devices to the Internet. While these standards define general solutions, problems such as mobility and adaptability to network and traffic dynamics have been left open.

In this thesis, we design and implement distributed QoS-aware communication protocols for low-power IoT networks. Our proposed methods can be easily integrated into the protocol stack designed by IETF for meeting QoS constraints of different applications. To find effective solutions, we model the problems of routing, congestion control, and Time-Slotted Channel Hopping (TSCH) by applying different optimization methods such as game theory, convex optimization, and fuzzy logic. To examine the performance of our proposed protocols in real industrial environments, we implement them on Zolerita Firefly motes and the Contiki operating system. The evaluation results show that our proposed methods can significantly improve the performance in terms of end-to-end delay, packet delivery ratio, and throughput in mobile and static networks.