Spirent and two leading UK universities deliver first-of-its-kind PNT data set

Cardiff University and the University of Bristol are global research leaders in the fields of robotics, autonomous navigation, and environmental sensing. At Cardiff University, Dr Ze Ji is a senior lecturer in Robotics and Autonomous Systems in the School of Engineering. He leads the university’s Robotics and Autonomous Intelligent Machines (RAIM) group and manages the Robotics and Autonomous Systems Laboratory, which provides cutting-edge research facilities including a range of robots and drones as well as vision systems and other sensors. At the University of Bristol, postdoctoral research fellow Dr Timothy Pelham is researching spatial intelligence as a new way of securing wireless networks. As part of his research, he has published Lycean EM; a framework for rapid virtual prototyping of antenna arrays and frequency and time domain channel modelling.

Both Dr Ji and Dr Pelham had identified a need for a navigation data set that included high-quality ‘ground truth’ recordings of the real-world global navigation satellite systems (GNSS) signal environment. Dr Ji is developing a localization framework that uses machine learning (ML)to enable autonomous systems to navigate reliably in GNSS or GNSS-denied environments. A data set that combined recordings of the live environment with synchronised readings from a GNSS receiver, cameras and inertial measurement units (IMUs), would be valuable both for lab simulations and for training ML algorithms for autonomous positioning, navigation and timing (PNT). Dr Pelham is working on algorithms for spatial fingerprinting, using a digital twin-based approach to merge the radio and spatial domains for secure positioning in adversarial environments. From his point of view, such a data set could help with developing applications for more secure acquisition of GNSS positions, including in the presence of interference, jammers, or other types of radio frequency (RF) noise.

Both academics had independently searched for a data set to meet their needs, but hadn’t found any that included a recording of the live GNSS environment alongside the other sensor measurements – so it was time to create one.