Publish, subscribe, and federate!

TL;DR

This paper proposes a scalable federation approach for multiple publish/subscribe brokers in IoT, enabling seamless data sharing among distributed brokers to improve system robustness and efficiency.

Contribution

It introduces a novel federation method for independent brokers in IoT publish/subscribe systems, addressing management and scalability challenges.

Findings

Enables subscribers to receive data from any broker regardless of publisher location.

Reduces bottlenecks and single points of failure in IoT broker networks.

Improves scalability and robustness of IoT publish/subscribe architectures.

Abstract

Connecting usual things/objects to the Internet allows the monitoring and control of such things from anywhere, which is usually referred to as the Internet of Things (IoT). Things communicate among themselves or with other entities (e.g., a server) so that information can be gathered from things whilst proper actions can be taken upon them. A prominent communication approach adopted by many IoT applications is related to the Publish/Subscribe (P/S) paradigm. Any communication entity willing to provide some data announces its intention to a server (broker), establishing itself as a publisher for such data/topic. Entities that are willing to receive any published data, register themselves to the broker as subscribers. While employing just one broker might lead to a bottleneck and a single point of failure, when having multiple brokers one could end up having difficulties with their management. This work presents a scalable and efficient proposal for the federation of independent brokers, by allowing subscribers to get all their publications no matter to which broker publishers and subscribers are associated with.