Speaker
Description
This talk will present some practical implementations of passive localisation techniques known as one-step algorithms. These methods are commonly reserved for offline, non-time-critical scenarios due to their reputation for poor scalability. However, in scenarios where they are applicable, they offer benefits such as improved robustness, particularly in challenging signal environments. Recent advances in low-cost, small form factor computing have made it increasingly feasible to incorporate computationally intensive optimisation techniques, including the one-step methods, into real-time systems without significantly degrading overall processing time.
We explore simplifying assumptions and operational conditions that enable real-time passive localisation using one-step algorithms. This includes field-tested results from a distributed localisation system using networked receivers and decentralised computation. We focus on airborne and ground-based sensors, and the corresponding data and synchronisation challenges encountered during system integration. We also discuss algorithmic techniques that have been successful for this application and that may be valuable to adjacent domains. GNU Radio has been used during the conceptualisation and lab-testing of various system components.
The presentation will be at a level accessible to a broad audience. This work includes ongoing collaboration between Flinders University and DSTG.