ABSTRACT

The group of real-world physical devices like machines, vehicles and various “things” connected to the Internet is called anInternet of Things (IoT). These things are deployed with certain hardware components such as microcontrollers/processors, sensors, actuators and networking capabilities along with certain amounts of programmability. These resources make these things smarter by associating with each other and exchanging the information themselves. The volume of the total number of IoT devices is expanding every year. It was 30 billion in 2020, and it is estimated that there willbe 60 billion gadgets by the year 2025. The IoT has widespread applications like smart city, smart grid, intelligent transportation, healthcare, etc. The major challenge in IoT is that it is fully dependent on the cloud for all kinds of computation, which leads to high latency in the IoT devices. To overcome this latency issue, the edge and fog computing applications are introduced. Edge computing, otherwise called merely ‘edge,’ processes data near the data source, and it does not send to remote cloud or other unified frameworks by which speed of execution is enhanced for processing the data. Fog computing is a standard that characterizes how edge processing should function, and it encourages the tasks of computing, stockpiling and computational systems. IoT protocols and applications do not take into consideration the fog computing, as IoT was conceived with cloud as a source for performing analytics. The MQTT protocol is one of the most widely accepted protocols in the IoT environment due to its simplicityand reliability features applied to embedded devices. The new upcoming changes in fog computing require new hardware changes, which increase the system complexity. To overcome these issues, a novel software-defined multi-tier fog computation model is proposed. Deep learning offers precise information when compared to the other learning algorithms. On the other hand, rapid growth in the area of Internet of Things (IoT) has been taking place, which has many application areas.