Target Turning Maneuver Detection using High Resolution Doppler Profile

Target maneuver discrimination using ISAR image in interception

“…The innovation-based [6] maneuver detector and state estimator depend on the innovation information of the Kalman filter or its variation. Whether a single-model or multiple-model method, it is difficult to achieve a short discrimination delay while maintaining high correct probabilities, due to the Q effect [2].…”

“…This inherent drawback becomes more serious when a mismatch occurs between target acceleration and predesigned models. By employing maneuver information embedded in the features from sensors, the feature-based technique [6] breaks through the above obstacle. In an air-to-air interception scenario, the bank angle measurement from the image is utilized in estimator when the fighter plane is taking a bank-to-turn (BTT) maneuver [4].…”

“…The radar echo also conveys maneuver information due to the modulation effects on electromagnetic scattering, and it can easily handle the problems of the optical sensors. Some narrowband radar features including glint, radar cross section (RCS), and high-resolution Doppler profile (HRDP) have been exploited successfully for type I maneuver detection [6,9,10]. But the MDS discrimination is not involved in the aforementioned researches.…”

“…The former methods depend on the innovation information of the Kalman filter or its variations [6]. Due to the Q effect [2], it is difficult for both of the single-model and multiple-model methods to achieve a short discrimination delay while maintaining high correct probabilities.…”

“…This inherent drawback becomes more serious with mismatch between target acceleration and predesigned models. However, employing maneuver information embedded in the features from sensors, feature-based techniques [6] overcome the above obstacle. For instance, in an air-to-air interception scenario where a fighter plane is taking a BTT maneuver [4], an estimator using the bank angle measurement from images can greatly reduce the detection delay, compared to the classical innovation-based estimator.…”