Smart Energy Harvesting for Internet of Things Networks

Area of Science:

• Computer Engineering
• Network Engineering
• Artificial Intelligence

Background:

• Internet of Things (IoT) nodes face limitations in battery life, impacting network performance and longevity.
• Femtocell access points (FAPs) can support IoT networks, but efficient energy management is crucial.
• Existing approaches often lack personalization, leading to suboptimal energy allocation and network efficiency.

Purpose of the Study:

• To introduce a smart energy harvesting framework for IoT networks to prolong battery life.
• To leverage Contract Theory and Reinforcement Learning for personalized energy management.
• To optimize the interaction between IoT nodes and FAPs for improved system-wide social welfare.

Main Methods:

• Classifying IoT nodes based on characteristics (IoT node types).
• Applying Contract Theory to design personalized contracts (transmission and charging power) between IoT nodes and FAPs.
• Formulating utility functions based on personalized profit for all entities.
• Utilizing Reinforcement Learning for autonomous IoT node association with beneficial FAPs.
• Conducting simulations to evaluate framework performance against alternative methods.

Main Results:

• The proposed framework enables personalized energy harvesting contracts between IoT nodes and FAPs.
• Reinforcement learning facilitates autonomous and optimal FAP association for IoT nodes.
• Personalized contracts demonstrated a four-fold improvement in IoT system social welfare compared to agnostic approaches.

Conclusions:

• The smart energy harvesting framework effectively prolongs IoT node battery life.
• Personalized contract design is a key factor in maximizing IoT network efficiency and social welfare.
• The integration of Contract Theory and Reinforcement Learning offers a robust solution for intelligent energy management in IoT networks.