Radar Error Calculation and Correction System Based on ADS-B and Business Intelligent Tools

With the growth of air transport, the air traffic control needs to enforce the Communication navigation surveillance air traffic management (CNS-ATM) because this is the back bone of the air operation in any country. This system has the responsibility of guaranteeing air safety and management of the national air space (NAS) that nowadays needs to increase the flight density to respond to the demand. To accomplish this, new technologies like air dependent surveillance broadcast (ADS-B) have been used to increase the accuracy and time response of data air surveillance sensor integration of sensor location and the reliability of ATM system. CNS-ATM system for surveillance and control of aircrafts have been mainly used in primary and secondary radars to calculate the aircraft position through signal delay or time difference between transponder pulses. The accuracy of each sensor depends on internal and external factors such as frequency, power, target distance, noise, maintenance, and others. When an aerodyne is detected by multiple sensors, it could create a multiple track in a geographic and temporal space where the aircraft will be possibly flying. This space depends of radar update time, aerodyne speed, and the accuracy of each sensor, and it is difficult to know where the aircraft really is. This work proposes a technique based on ADS-B for making an error calculation of each sensor in a fusion system, using business intelligence techniques for understanding the error condition of each sensor in a geographical area. Based on results, we propose a technique that could make an error correction to avoid phase shifts between sensors. The information of this data study was used for statistical calculation values such as variance and standard deviation. For fusion accuracy improvement, three steps have been proposed in this research. First, the use of the radar error by region and statistical values by calculating the Kalman filters for each sensor to reduce the internal error of the radar. Second, the bias measured against ADS-B signal, used like a parameter to calculate radar bias correction that could be applied as a feedback input in a homogenization signal process or tracking process to reduce sensor bias in a recurrent process. Third, the use of Kalman prediction characteristic to replace missing points in a trajectory calculation. This technique was implemented by Colombian system to reduce error and bias sensor and a user's quality perception in a radar tracking and fusion track system in a surveillance network. In this process, it was found that it is possible to use it by a repetitive error measured ADS-B track like a reference track to calculate the error and in this way, it could be possible to reduce the uncertainty about the aircraft position. On the other hand, the use of data analysis process based on business intelligent tools allows us to easier understand the radar error behavior. Both methodology and results will be described here.