Models for automatic aircraft type prediction

Abstract

Air defense is one of the major missions of the Air Force. The first step is to find objects with speed which could pose a threat to national security by using radar. Basically, any object first detected by radar is determined to be an unknown target. In other word, it is a flying object which type and owner have not yet been identified. An accurate analysis of aircraft type and owner is an important factor for the decision on how the RTAF responds to the situation. For example, if it’s an opponent's combat aircraft, the RTAF will be able to intercept and alert aircraft units to defend the air base with appropriate tactics. At present, the analysis and identification of unknown targets are being handled by expert staffs with experience in operating and training with allied forces. The accuracy of the analysis depends on the capabilities of each officer. Training the staff to be competent and specialized in this field takes several years. The capability is also relevant to the aviation services that external agencies require, where the compensation is much higher. As a result, the resignation of the personnel of this group to sign contract with such agencies continuously occurs. To solve this issue, the researcher constructed a model for predicting unknown aircraft using the data from the air traffic analysis system of neighboring countries owned by Directorate of Intelligence by implementing data classification technique i.e. the Bayesian Network, SVM, and Decision Tree. The best result was a model derived from the Random Forest technique, because of its high accuracy in prediction. The researcher believes that this model can improve the decision-making process of the commanders more quickly and appropriately. The results of this research will be able to solve problems of expert personnel shortage and reduce the dependence on the expertise of specialized staffs in accordance with the concept of Smart Weapon Systems, which lead to the selection of the Smart Tactics of the RTAF.