Experimental RCS acquisition system: Using software defined radio to build a classification dataset

One of the major issues in testing Automatic Target Recognition (ATR) systems lies in having a sufficient amount of trusted data to be able to estimate efficiency and reliability of the developed algorithms. Generally synthetic data is used because real world data is not easy to obtain or too expensive for academic research. This paper describes an original low-cost automatic acquisition system designed to build a database of RCS acquired with real planes (airliners), in bistatic configuration, at low frequency (VHF). The described system uses ADS-B for target position and a network connected set of Software Defined Radio (SDR) receivers for RF processing. This setup was used to build a RCS database of 1300 commercial airliners in less than two weeks close to Paris-Orly international airport. The web-based graphical user interface was used over this time to remotely monitor the system and make demonstrations.