ABSTRACT
The escalation in connected Internet of Things (IoT) devices has created a huge demand for artificial intelligence (AI)-based methodologies. Though revolutionary, IoT is associated with limited computational capabilities in terms of storage and processing. IoT alone also suffers from certain issues such as privacy, reliability, security, and performance. Hence, IoT requires integration with the Cloud of Things (CoT), which paves the path to resolve many of these issues. With the exponential growth of IoT applications, traditional centralized cloud computing faces challenges such as limited capacity, network crashes, high latency, and many more. Because the number of devices in smart environments is on rise, the response time required by these devices requires IoT data to be managed and processed near the same location where it is being generated (i.e., at the edge). Fog computing too complements IoT and edge computing by providing data storage and data processing at IoT devices instead of directing the data to the cloud. Hence, such IoT-based architectures are required to preserve the benefits of both IoT-enabled cloud computing, edge processing, and fog processing. AI and its supplements provide support in achieving IoT-based environments. However, most AI and machine-learning algorithms require a substantial amount of processing power, which may or may not be available at the edge. Current researchers hope to develop some architecture that strikes a positive balance between all the benefits and requirements of IoT and edge computing. This chapter includes an introduction to IoT and AI, IoT and edge computing, and IoT and fog computing, and their various building blocks. Discussion of the benefits and security aspects of edge and fog computing is also included.
