[Paper Review] A Review on Recent Energy Harvesting Methods for Increasing Battery Efficiency in WBANs

Hossein Yektamoghadam, Amirhossein Nikoofard+2 more|arXiv (Cornell University)|Jan 12, 2024

Energy Harvesting in Wireless Networks4 citations

TL;DR

This paper reviews recent energy harvesting techniques—thermal, kinetic, RF, and biofuel cell-based methods—for extending battery life in wireless body area networks (WBANs), particularly in implantable biomedical devices. It proposes integrating reinforcement learning and distributed optimization to enhance energy efficiency, offering a roadmap for next-generation rechargeable WBAN sensors.

ABSTRACT

Today, technology development has led humans to employ wearable and implantable devices for biomedical applications. An important research issue in this field is the wireless body area networks (WBANs), which focus on such devices. In WBAN, using batteries as the only energy supply is a significant challenge, especially in medical applications. Charging the batteries is a problem for patients who use WBAN. Replacing the battery is not very difficult for wearable devices, but implantable devices have different conditions. The use of batteries in implantable devices has many problems, including pain and costs due to surgery, mental stress, and lack of comfort. Batteries' life depends on their type, operation, the patient's medical condition, and other factors. This paper reviews recent energy harvesting methods for battery recharge in WBAN's sensors. Moreover, we provide future research directions on energy harvesting methods in WBANs. Therefore, active research fields such as reinforcement learning (RL) and distributed optimization in WBAN applications were investigated. We strongly believe that these insights will aid in studying and developing a new generation of rechargeable sensors in WBANs for fellow researchers.

Motivation & Objective

• Address the critical challenge of limited battery life in implantable and wearable WBAN devices.
• Overcome the limitations of battery replacement in implantable devices, which require invasive surgery and cause patient discomfort.
• Explore and evaluate recent energy harvesting technologies to enable sustainable, long-term operation of WBAN sensors.
• Identify future research directions in energy harvesting, particularly leveraging reinforcement learning and distributed optimization for dynamic energy management.
• Provide a comprehensive review to guide researchers in developing next-generation rechargeable WBAN systems with improved energy autonomy.

Proposed method

• Systematically reviewed energy harvesting sources relevant to WBANs, including thermal gradients from the human body, kinetic motion, radio frequency (RF) signals, and biofuel cells.
• Evaluated the technical feasibility, power output, and integration challenges of each harvesting method in wearable and implantable contexts.
• Analyzed the role of reinforcement learning (RL) in optimizing energy collection and allocation based on real-time physiological and environmental data.
• Explored distributed optimization techniques to coordinate energy harvesting and consumption across multiple sensors in a WBAN.
• Synthesized findings into a framework for intelligent energy management, emphasizing adaptability and scalability in dynamic WBAN environments.
• Provided a comparative analysis of harvesting methods using performance metrics such as power density, reliability, and biocompatibility.

Experimental results

Research questions

• RQ1 What are the most viable energy harvesting sources for extending battery life in WBAN sensors, especially in implantable applications?
• RQ2 How can reinforcement learning be applied to dynamically optimize energy harvesting and consumption in WBANs?
• RQ3 What are the key technical and biological constraints affecting the integration of energy harvesting in WBANs?
• RQ4 How do distributed optimization techniques improve energy efficiency across multi-sensor WBANs?
• RQ5 What future research directions are most promising for enabling fully autonomous, rechargeable WBAN systems?

Key findings

• Thermal energy harvesting from body heat offers a continuous but low-power source, suitable for low-data-rate sensors.
• Kinetic energy harvesters, such as piezoelectric and electromagnetic devices, show promise for motion-activated applications but face challenges in consistent power output.
• RF energy harvesting enables battery-free operation in proximity to RF sources but is limited by low power density and signal variability.
• Biofuel cells using glucose and lactate as fuel sources demonstrate high theoretical energy density and biocompatibility, though practical implementation remains challenging.
• Reinforcement learning enables adaptive energy management by learning optimal harvesting and duty-cycling strategies based on real-time data and environmental conditions.
• Distributed optimization techniques enhance system-wide energy efficiency by coordinating energy harvesting and data transmission across multiple nodes in a WBAN.