Advancements in Machine Learning-Enhanced Green Wireless Sensor Networks: A Comprehensive Survey on Energy Efficiency, Network Performance, and Future Directions

Wireless sensor networks (WSNs) are a collection of sensor nodes that collect data from the environment using wireless technology. WSNs have many applications in various domains, such as public utilities, industrial monitoring and control, and defense and military activities. However, WSNs have limited energy, short network lifetime, high bandwidth requirements, low throughput (TP), and unreliable connections. Green WSNs (GWSNs) are approaches that optimize energy consumption and enhance sustainable networks. Despite these advancements, nonadaptability to dynamic network conditions and the use of static historical data necessitates introducing machine learning (ML) techniques to address these challenges. GWSNs aim to reduce energy consumption and environmental impact, while ML techniques will improve data processing and network performance. This paper surveys recent advances in ML-based GWSNs, covering different aspects such as network structure, data exchange, location information, quality of service (QoS), and multiple path support. We also present the performance metrics, implementation issues, and future trends in ML-based GWSNs. The paper introduces a new taxonomy categorizing ML applications based on network architecture, data sharing, location data, multipath support, and QoS. The survey findings show that ML-based GWSNs can achieve up to 50% energy savings, 30% TP improvement, and a 40% delay reduction (DR) compared to conventional WSNs.