Energy-efficient and reliable data transfer in practical bluetooth low energy-based wireless sensor networks

Abstract

Wireless Sensor Networks (WSNs) are pivotal in high-sampling rate applications such as industrial vibration monitoring, structural health monitoring (SHM), and motion sensing in healthcare and livestock management. These applications often gather accelerometer data at frequencies ranging from 10 Hz to several hundred Hz, necessitating rapid and energy-efficient data transmission. Focusing on many-to-one WSN topologies, the thesis centers on the use of Bluetooth Low Energy (BLE) advertising schemes, and develops reliable, energy-efficient data transfer mechanisms for such data-intensive scenarios. While previous research has predominantly explored BLE legacy advertisements for real-time data transfer from the transmitter’s perspective—often neglecting the practical challenges of transmitting dynamic sensor data where high data loss can not be tolerated since the data represents the current state of the system— this thesis specifically addresses these issues and proposes a mechanism to overcome them. In addition, we also examine extended and periodic advertising schemes, which have never been studied in the existing literature for data transfer, even though they were introduced by the Bluetooth Core Specification with enhanced features to support data transfer in WSN deployments. A detailed evaluation of the three distinct schemes is presented, accompanied by recommendations and best practices for their implementation in real-world WSN deployments. This study also highlights some application-specific challenges that may be encountered while using commercially available chipsets that strictly adhere to the BLE link layer functions as defined in the Bluetooth Core Specification. The analysis highlights the benefits and limitations of each advertising method, noting that different chipset capabilities may restrict developers’ options. Furthermore, the thesis introduces energy-efficient relay-based data propagation strategies for WSNs based on all three types of advertising schemes. Additionally, a novel BLE gateway architecture is developed, implemented, and rigorously evaluated; this connectionless gateway design minimizes processing overhead and data loss at the application, representing, to the best of our knowledge, the first practical realization of such a system. Broadly, the research offers a comprehensive assessment of BLE advertising-based data transfer mechanisms for WSNs and proposes strategies to enhance their energy efficiency, demonstrating their potential to meet the demanding requirements of high-data-rate IoT applications while maintaining low power consumption.