Description of field of research:

UWB is an emerging technology that has drawn much attention in recent years. IR-UWB devices can generate very narrow electromagnetic pulses which can be utilized in precise positioning and range measurement. 

Our group has developed a few UWB-based applications including pedestrian tracking & navigation, collision avoidance. 

However, the UWB technology requires a sub-nanosecond level accuracy in time-of-flight measurement, which makes it sensitive to the Non-Line-Of-Sight(NLOS) scenario. 

The knowledge of NLOS can help to improve the precision of the UWB-based applications.

This project is about to recognize the NLOS scenario by analyzing the channel impulse response (CIR) data. Machine learning might be used to process the CIR data.

School

Minerals and Energy Resources Engineering

Research areas

Embedded Systems, Machine Learning

The candidate will work closely with Dr Binghao Li, Kai Zhao, Dr Boge Liu and other team members of the MIoT & IPIN lab. Dr Binghao Li is the main contact.

  • The candidate is expected to set up a machine learning training dataset by collecting data with different devices in different scenarios. In order to automate the data collection, the candidate may also need to develop some small tools/scripts.
  • It is also expected that the dataset can be processed with some machine learning models and the performance of the models in NLOS recognition can be compared and analyzed.

The target chipset:

  • https://www.decawave.com/product/dw1000-radio-ic/

The description of the proposed issue:

  • https://www.decawave.com/wp-content/uploads/2018/10/APS006_Part-3-DW1000-Diagnostics-for-NLOS-Channels_v1.1.pdf

An example of our UWB application:

  • Li, B., Zhao, K. and Sandoval, E.B., 2020. A UWB-Based Indoor Positioning System Employing Neural Networks. Journal of Geovisualization and Spatial Analysis, 4(2), pp.1-9.