MQTT2EdgePeer: a robust and scalable peer-to-peer edge communication infrastructure for topic-based publish/subscribe

Abstract

In recent years, there has been an unprecedented growth in the number of Internet of Things (IoT) devices, resulting in a widespread adoption of interconnected devices across various sectors such as residential, medical, transportation, agriculture, and industrial domains. This surge in IoT devices has brought forth both opportunities and challenges, necessitating careful consideration of communication aspects in the design of IoT protocols and systems. One critical factor to consider in the design of IoT communication is the need for efficient bandwidth utilization. IoT devices typically have limited bandwidth resources, which must be carefully managed to accommodate the increasing volume of data generated and transmitted by these devices. Another crucial consideration in IoT communication design is the importance of low latency. Many IoT applications require real-time or near real-time data exchange for timely decision-making and responsiveness. Therefore, minimizing latency, or the time delay in data transmission, becomes essential to meet the stringent requirements of latency-sensitive IoT applications. In this study, we introduce MQTT2EdgePeer, a novel topic-based publish/subscribe system constructed upon a structured peer-to-peer overlay network for efficient dissemination of messages. The dissemination of messages is facilitated through both single-hop and multi-hop coordinator-based approaches. MQTT2EdgePeer seamlessly integrates with the standard MQTT protocol, allowing effortless connection for any IoT applications. The implementation and deployment of MQTT2EdgePeer at the edge are presented, accompanied by experimental evaluations that include a comparative analysis with a traditional single rooted tree approach. The results obtained validate the efficacy of MQTT2EdgePeer in various aspects, including load distribution, latency, bandwidth usage, scalability, and fault-tolerance.

Date	31 Jul 2023
Original language	American English
Awarding Institution	École de technologie supérieure
Supervisor	Julien Gascon-Samson (Supervisor)