Integration of artificial intelligence (AI) with sensor networks: Trends, challenges, and future directions

Abstract

Wireless sensor networks (WSNs) have become widely ubiquitous deployed in many application domains over the past few decades. Classical approaches configure WSNs statically which makes altering and re-configuring them dynamically a challenging task. To address this challenge, Artificial Intelligence (AI) technologies could be integrated into WSNs. The utilization of AI systems holds the potential to enhance the efficient management of energy consumption, thereby preventing wastage and ensuring prolonged usage in WSN. This research investigates the application of AI, encompassing intelligent computing techniques such as Machine Learning, to promote sustainable energy practices, with a particular focus on Internet of Things (IoT) devices. We explores various methodologies employed in AI, including supervised and unsupervised learning, as well as reinforcement learning, within the context of existing practices. The primary objective is to provide insights for researchers seeking to comprehend recent advancements in employing AI for sustainable energy of WSNs. Additionally, the article addresses persistent challenges and issues that require attention and optimal solutions in this domain. As such, we provide an in-depth analysis of various recently proposed AI strategies (2007 to 2023) that have been applied to WSNs. The examined articles have been systematically arranged to enhance follow-through, clarity, and readability. Furthermore, we discuss the benefits and drawbacks of the (AI) techniques employed in WSNs. Finally, we discuss future research opportunities that would enable scalable and cost-effective AI deployment in WSNs.

Keywords

• Artificial intelligence
• Deep learning
• Energy-efficiency
• Machine learning
• Wireless sensor networks