Publish/subscribe based multi-tier edge computational model in Internet of Things for latency reduction

Most of the Internet of Things (IoT) applications are time sensitive and require low latency, as a millisecond delay can affect a huge production in the machine-dependent environment. Due to the rapid growth of the internet connected devices, it is very difficult to handle this huge processing load in the Cloud and traffic in the network. Data communication protocols in vogue are based on IoT devices connected to Cloud. But with the introduction of edge/Fog computing in IoT where all processing happens at the edge, there is a need to revisit the existing IoT protocols. The existing protocols were not designed with edge/Fog computing in mind. The most predominantly used IoT protocol is Message Queue Telemetry Transport (MQTT) protocol which is based on publish subscribe model. This paper proposes software defined multi-tier edge computing model by modifying the existing MQTT protocol for edge computing with a remote broker in Fog node and main broker in Cloud. A mathematical model is proposed towards computing the performance metrics of the proposed system. The performance of the proposed system is been compared with the traditional MQTT based IoT model. Python-Virtual Fog (PVFOG) simulator has been used to measure the performance in terms of service latency, transmission latency, processing latency, and packet send to Cloud. The experimental results show that the proposed system outperformed the traditional MQTT based IoT Model. (C) 2019 Elsevier Inc. All rights reserved.