A Model-Driven Approach for Load-Balanced MQTT Protocol in Internet of Things (IoT)

Abstract

Internet of Things (IoT) has changed view of people about devices and has paved new ways of interaction. These devices form a powerful network that can be controlled from anywhere anytime. Though this is beauty of IoT but it comes with high price of vulnerability and high load. Broker based IoT protocols enable two way communication. One of these protocols is Message Querying Telemetry Transport (MQTT). Load balancing in IoT is huge problem when it comes to application of such publish subscribe models. This issue provides loophole for adversaries to exploit by attempting various attacks on network e.g. Denial of Service (DoS), Distributed Denial of Service (DDoS) etc. These attacks not only deprive network of its bandwidth but also are threat to confidentiality and privacy of any IoT based system. In this paper, we have tried to investigate load balancing problem by investigating different MQTT based threat models and proposing a UML profile for effective handling of load balancing in IoTs employing MQTT protocol. Intuition of proposed profile is to introduce lightweight extension that can provide a robust load-balanced version of MQTT protocol. Validation of profile is performed through case study of Proactive Technical Support System for HealthCare. From validation it is evident that proposed profile is robust in handling load balancing shortcomings in IoT ecosystems.