Research On Confidence Of Wide Area Multilateration Based On ADS-B Signal

Accuracy and reliability of localization and tracking of aircraft have great influence on their safety in the system of air traffic control(ATC), thus researches on two kinds of surveillance technology, automatic dependent surveillance-broadcast(ADS-B) and wide area multilateration(WAM), have great application value. The error sources, localization and tracking accuracy of both monitor sensors, ADS-B and WAM, were analyzed in this paper. Besides, simulation experiments were utilized to analyze the average Mahalanobis distances between ADS-B position reports and predicted locations as well as the root mean square(RMS) of position errors of associated results under different ADS-B latency, to evaluate tracking effectiveness of the nearest neighbor data association(NNDA) and the probability data association(PDA).Firstly, WAM need utilize time of arrival(TOA) of target signals to position. Error sources of TOAs include system errors caused by hardware measurement and random errors. Random errors include: measurement errors caused by noise; measurement errors caused by sampling; synchronization errors of global positioning system(GPS). In addition, ADS-B position report errors include: GPS position errors; quantization errors of ADS-B position; ADS-B position errors caused by latency.Secondly, it was proved that assuming the variances of the TOA measurement errors of four base stations are different, variances of WAM position errors equal weighted sums of all variances of TOA measurement errors in this paper. Then assuming that all base stations have the same variance of the TOA measurement errors and common synchronization error is absent, it has been demonstrated that TOA locations outperform TDOA locations in performance based on geometric dilution of precision(GDOP).Finally, simulation experiments were utilized to calculate average Mahalanobis distances between ADS-B position reports and predicted locations of tracking, as well as, RMS of position errors of associated results under different ADS-B latency, to compare two data association methods. Analytical results indicate that PDA methods outperform NNDA methods in tracking error and robustness in the presence of ADS-B latency.