A Distributed MQTT Broker System for Location-based IoT Applications
Ryo Kawaguchi, Masaki Bandai
Abstract
Ryo Kawaguchi, Masaki Bandai
Abstract
Message queue telemetry transport (MQTT) is well known as an application layer communication protocol for the Internet of Things (IoT). Since basic MQTT defines one broker in a system, the number of publishers and subscribers is limited. Therefore, there are MQTT implementations that perform load balancing via cooperation with multiple brokers. However, when subscriptions frequently change for mobile clients, as is the case when handling location-dependent data, or when the number of clients is large, information sharing between brokers increases. In addition, with existing implementations, different kinds of broker cannot cooperate. In this paper, we propose a distributed broker system for large-scale location-based IoT services. The proposed method introduces a topic structure suitable for handling location-dependent data. In addition, distributed brokers and gateways are introduced to reduce broker load and to support heterogeneous brokers in a system. A prototype implementation and a theoretical evaluation are used to show the effectiveness of the proposed system.
OpenAlex reports 24 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Message queue telemetry transport (MQTT) is well known as an application layer communication protocol for the Internet of Things (IoT). Since basic MQTT defines one broker in a system, the number of publishers and subscribers is limited. Therefore, there are MQTT implementations that perform load balancing via cooperation with multiple brokers. However, when subscriptions frequently change for mobile clients, as is the case when handling location-dependent data, or when the number of clients is large, information sharing between brokers increases. In addition, with existing implementations, different kinds of broker cannot cooperate. In this paper, we propose a distributed broker system for large-scale location-based IoT services. The proposed method introduces a topic structure suitable for handling location-dependent data. In addition, distributed brokers and gateways are introduced to reduce broker load and to support heterogeneous brokers in a system. A prototype implementation and a theoretical evaluation are used to show the effectiveness of the proposed system.
Key concepts: MQTT, Message queue, Computer science, Implementation, Computer network, Distributed computing, Internet of Things, Queue