Improving UWB Indoor Localization Accuracy Using Sparse Fingerprinting and Transfer Learning

Indoor localization systems become more and more popular. Several technologies are intensively studied with application to high precision object localization in such environments. Ultra-wideband (UWB) is one of the most promising, as it combines relatively low cost and high localization accuracy, especially compared to Beacon or WiFi. Nevertheless, we noticed that leading UWB systems’ accuracy is far below values declared in the documentation. To improve it, we proposed a transfer learning approach, which combines high localization accuracy with low fingerprinting complexity. We perform very precise fingerprinting in a controlled environment to learn the neural network. When the system is deployed in a new localization, full fingerprinting is not necessary. We demonstrate that thanks to the transfer learning, high localization accuracy can be maintained when only 7% of fingerprinting samples from a new localization are used to update the neural network, which is very important in practical applications. It is also worth noticing that our approach can be easily extended to other localization technologies.