OTDOA/GPS Fusion for Urban UAS Navigation using Particle Filtering Techniques

As Unmanned Aircraft Systems (UAS) become more accepted into our society, law enforcement, emergency services, and commercial entities will begin taking advantage of this technology in high-population urban settings. Since these aircraft will be operating between and above the high-rise buildings that de ne most urban cores, increased position estimation accuracy will become paramount to ensure safe and reliable operations. Fighting against this need for better accuracy is the degradation of Global Positioning System (GPS) accuracy in urban canyons due to signal re ection and refraction caused by buildings and other tall objects. UAS cannot rely solely on GPS position measurements for guidance and navigation to e ectively conduct urban missions. A potential solution to achieve the necessary positioning accuracy is through the fusion of the GPS data with cellular Observed Time Di erence of Arrival (OTDOA) position measurements. In areas with even modest GPS coverage, today's smart phones are already equipped to provide both measurements thus could plug into a UAS to serve as the low-cost urban autopilot of the future. Because OTDOA data is noisy, data ltering and fusion is essential. This paper proposes a sampling importance resampling particle lter to estimate UAS position from GPS and OTDOA data. Expected data error statistics are summarized from the literature and used in simulation to evaluate positioning accuracies using GPS alone, OTDOA alone, and a fusion of both data sources. Results show the GPS augmented by OTDOA provides a more accurate estimate of UAS position than either alone. However, OTDOA measurements from cellular networks of today are not as accurate as GPS measurements, thus are not yet a good sole source of aircraft position measurements, suggesting augmentation by another secondary sensing source distinct from GPS in GPS-denied areas in future work.