Recent progress in designing tiny wireless sensors have enabled the development of Wireless Body Area Networks (WBAN). A WBAN is composed of a few or tens of miniaturized wearable sensors which wirelessly communicate with a gateway. WBAN supports a wide range of applications from physical activity detection and motion tracking to ambulatory health monitoring.
Existing communication protocols developed for WBANs commonly adopt fixed-power transmission, in which the transmission power for sending packets from the sensor node to the gateway at different times is the same regardless of channel quality fluctuations. Such schemes cannot achieve good performance in terms of energy consumption, interference, and communication reliability.
In this talk, I will first give a brief introduction of WBANs and the challenges for communication in WBANs. Then I will present a light-weight scheme that enables sensors to self-learn the channel quality fluctuations using the locally measured inertial data. Using this algorithm, the sensor node can estimate the channel on a real-time basis and adjusts its transmission power at a per packet level. This scheme is validated through implementing the idea on the TelosB platform.
In the second part, I will show how location information can help with channel quality estimation. I will first discuss how to develop a 3D body motion tracking system using 9-axis Inertial Motion Units (IMU) and give a demo. Motivated by the idea of motion tracking, I will present a new localization method in which the sensor can estimate its location relative to the gateway simply based on motion data and the bio-mechanical constraints.
Last modified: Tuesday, 24-Apr-2018 10:57:31 NZST
This page is maintained by the seminar list administrator.