Analytical review of artificial intelligence applications in improving security and energy efficiency in wireless sensor networks

Due to the rising application of Wireless Sensor Networks within the context of the Internet of Things, environment surveillance, intelligent industries, and healthcare, the availability of strong security along with low energy resources is an inherent challenge faced in the design process of Wireless Sensor Networks. On the one hand, the processing capabilities, as well as the energy resources, are meager in sensor nodes, while, on the other hand, the concerns of security, threats, and vulnerabilities create significant impacts on the efficiency of these sensor networks. During the last decade, Artificial Intelligence with areas including Machine Learning, Deep Learning, Evolutionary Algorithms, and Reinforcement Learning, has emerged with promising attributes to improve the efficiency, security, and energy resources of Wireless Sensor Networks. In the course of my thesis, we will systematically review the latest works on the application of AI-based solutions to enhance the security and energy efficiency aspects of WSN technology, from the year ۱۹ until the year Y. Yε. To start, the basic challenges in the design of these sensor networks are presented, followed by the intelligent application of AI solutions, systematically classified on the two axes: 'security' and 'energy efficiency.' The main outcomes show the efficiency of the combination of Fed-Learning, Edge-Learning, and TinyML solutions in defining the future course of intelligent self-adaptive sensor network development.

Keywords:

Wireless Sensor Networks (WSN) , Artificial Intelligence (AI) , Network Security , Energy Optimization , Reinforcement Learning , Federated Learning