Knowledge-based Defect Prediction and Test Case Generation for Radar Software Fault Prevention

Software is playing an increasingly important role in modern radar software. Software faults often lead to serious system failure. A fault prevention approach is proposed for radar software. The proposed approach follows the quality assurance link of “knowledge-prediction-control” and combines knowledge collection, defect prediction, test case recommendation, and automatic test case generation. Case studies show the proposed approach can effectively help fault detection on radar software.