<title>Real-time modeling of a radar sensor for enhanced vision</title>

Abstract

Millimetric radar imaging systems have been used to improve situational awareness for flight crew in low visibility approaches. The image from the sensor is projected in a Head-Up Display (HUD) and, for aircraft without Cat-III auto-land facilities, can provide sufficient cues to continue a manual approach past the normal decision height. However, these images may be cluttered, features are often difficult to detect and there is no direct indication of system integrity. Guidance cues can be displayed in the HUD by tracking runway features in the radar image and the use of sensor fusion methods to detect variation in the size and shape of the runway can provide integrity monitoring, for timely warning of system malfunction. In order to develop real-time tracking algorithms, it is necessary to generate synthetic radar images, which exhibit the properties of actual millimetric radar sensors. This paper outlines the model of a radar sensor used to generate real-time radar images incorporating appropriate attenuation and clutter properties. These images are derived from standard 3D visual databases and have been integrated in a flight simulator using a commercial image generation system. The radar model incorporates the effects of material properties of objects, the sensor range and grazing angles and includes atmospheric attenuation. Examples of the radar images are presented in the paper together with a summary of the real-time performance of the radar model to simulate millimeter wave radar images using a proprietary workstation.