Improving registration correction accuracy via finer quantization and timestamp

In order to successfully fuse measurements from multiple surveillance sensors in the National Airspace System (NAS), radar registration techniques are used to minimize radar biases in the measured range and azimuth. These biases can be systematically estimated; the estimated biases can then be subtracted from each target's measured range and azimuth to improve the measurement accuracy. The NAS's current radar's range and azimuth quantization, however, through hardware known as the Common Digitizer, has large quantization errors, potentially degrading registration estimation accuracy. The finer quantization of All Purpose Structured Eurocontrol Surveillance Information Exchange (ASTERIX) would theoretically improve the accuracy of registration bias estimates. In addition to more precise range and azimuth quantization, ASTERIX formatted data is required to include a Universal Coordinated Time (UTC) timestamp, eliminating the errors in the current system's time-of-applicability (TOA) estimation. In particular, errors stem from time alignment message estimation and variable sweep rates. Using PC-based simulations of long range radar sweeps, aircraft targets, radar noise, and registration biases, the authors propose to measure the net effect of finer range and azimuth quantization and more exact timestamps. Radar registration techniques used by US terminal and en route surveillance systems - those used by the Standard Terminal Automation Replacement System (STARS) and En Route Automation Modernization (ERAM)-will be faithfully replicated and simulated 2D aircraft flight patterns will be incorporated.