A data-driven approach to increasing the lifetime of IoT sensor nodes

Area of Science:

• Computer Science
• Electrical Engineering
• Agricultural Technology

Background:

• Wireless sensor networks (WSNs) face significant energy drain from data transmission, limiting applications like large-scale livestock monitoring.
• Efficient data handling is crucial for extending the operational lifespan of WSNs in remote or resource-constrained environments.

Purpose of the Study:

• To introduce Ambrosia, a novel lightweight protocol designed to reduce data transmission energy consumption in WSNs.
• To evaluate the effectiveness of Ambrosia in a real-world livestock monitoring scenario, focusing on data reduction and energy efficiency.

Main Methods:

• Development of Ambrosia, a protocol employing a window-based time-series forecasting mechanism for data reduction.
• Implementation of a configurable error threshold within Ambrosia to maintain data accuracy for end applications.
• Experimental validation using LoRa and Bluetooth Low Energy (BLE) on a livestock monitoring testbed.

Main Results:

• Achieved a 60% reduction in data transmission volume.
• Demonstrated a twofold (2x) increase in battery lifetime for the sensor nodes.
• Verified that the data reduction did not compromise the accuracy of the end application.

Conclusions:

• Ambrosia offers a viable solution for energy-efficient data transmission in WSNs, particularly for livestock monitoring.
• The protocol's data reduction capabilities directly translate to extended device operational life and reduced maintenance.
• Time-series forecasting presents a promising approach for optimizing data handling in resource-limited wireless sensing applications.