Blog
Why Standard Kalman Filters Lose Accuracy During Platform Manoeuvres: How Adaptive Noise Covariance Estimation Corrects It
A Kalman filter tuned for steady-state conditions will degrade the moment its host platform manoeuvres. Adaptive noise covariance estimation, driven by fuzzy logic, maintains tracking accuracy across the full operational envelope without any manual retuning.
Multi-Platform Target Tracking: Why Naïve Sensor Fusion Produces Inconsistent Results
In a decentralised sensor network, cross-correlations between platform tracking estimates are unknown and cannot be recovered. Standard Kalman fusion ignores this and overstates certainty. Covariance Intersection provides a rigorous, consistent fused track without that information.
Root-MUSIC Direction Finding: ±3-Degree Accuracy from a Four-Element Array
Root-MUSIC exploits the noise subspace of the array covariance matrix to extract precise angle-of-arrival from only four antenna elements. Combined with Kalman filtering, it maintains bearing accuracy to ±3 degrees without additional receiver hardware channels.
Can COTS Hardware Deliver Accurate RF Direction Finding Within SWaP Constraints?
A direction finding system built entirely from COTS SDR hardware achieved ±3-degree bearing accuracy at a 100-metre range and fitted within a single field-deployable Peli case. Rigorous algorithm design, not hardware cost, determines the result.
Integrating Mathematical and Computational Techniques to Solve Complex Engineering Problems
So, I have a new problem to solve! I have thought a little about it and am now sitting at my desk, with my extending pencil and some sheets of A4 in landscape orientation, writing