A Comparative Analysis of Challenges in Wireless Sensor Networks using Machine Learning Algorithms

Artificial intelligence (AI) and machine learning (ML) exhibit effective performance for the management of automated operations in wireless sensor networks (WSN). The WSN technology is broadly utilized in different applications such as health monitoring, environmental sensing, waste management, etc. WSN comprises tiny nodes utilized for effective data transmission towards the remote location using gateways based on wireless standard IEEE 802.15.4. The WSN is subjected to different challenges regarding network coverage, connectivity, energy consumption, bandwidth, security, maximization of network lifetime, and communication protocols. To overcome those issues, a WSN intrusion detection system (IDS) was implemented based on ML; this paper reviewed the contribution of ML to the WSN to overcome the challenges. Through the review of diverse articles, it is observed that power consumption is considered the major focused issue. The review shows that the supervised ML approach is broadly applied to overcome the issues in WSN. In 60 review articles, 63 percent focused on supervised learning, unsupervised learning accounted for 13 percent, and reinforcement learning accounted for 24 percent.