From Cloud to Edge: Collaborate MQTT Brokers as a Service
Video
Team Information
Team Members
Yanwen Jing, MS Student, Department of Electrical Engineering, School of Engineering & Applied Sciences (SEAS), Columbia Engineering
Luoyao Hao, PhD Candidate, Department of Computer Science, School of Engineering & Applied Sciences (SEAS), Columbia Engineering
Faculty Advisor: Henning Schulzrinne, Professor, Department of Computer Science and Department of Electrical Engineering, Columbia University
Abstract
MQTT (Message Queuing Telemetry Transport) is one of the de-facto application level protocols for IoT systems, which relies on a centralized broker on the cloud to forward messages from publishers to subscribers. With the increasing number of IoT devices joining services, traditional cloud-based broker deployment shows its deficiency in scalability. Moving brokers to edge computational nodes therefore comes into our vision. In this poster, we investigate the model and mechanisms to coordinate and collaborate distributed brokers towards an identical service. We propose an edge-assisted collaboration model with multiple clusters of brokers deployed on edges. With a focus on inter-cluster and intra-cluster communications, algorithms of loop-free, leader selection, and message dissemination are well considered. Extensive evaluations are expected to validate our approach in the near future.
Contact this Team
Team Contact: Yanwen Jing (use form to send email)
