Using code based GNSS double differences as Beacon landing system

Abstract

Due to time constraints this paper unfortunately only contains a fraction of the analysis and data collected during flight trials. For a detailed comparison see the future journal publication of this research. We report on using Global Navigation Satellite System (GNSS) pseudorange based double differences for relative navigation in the form of a beacon landing system and present an associated display and results of flight test data. In surveying, real time kinematics is an established technique that uses GNSS carrier phase measurements for highly precise position estimation. Part of the RTK process is to form the double difference observable between master and rover receiver and between pairs of satellites. With this double difference and knowledge of the satellite geometry, it is possible to calculate a relative position between two GNSS receivers. When using carrier phase measurements, the rover needs to be stationary for some time in order to initialize the resolution of the carrier phase ambiguities. The same technique can be used on pseudorange measurements without the limit of stationarity for initialization, albeit at the cost of a decreased accuracy. Almost all correlated errors are removed from the solution in the process of differencing [1] and the resulting DD observable has an accuracy that is comparable to the ground based augmentation system. For the double difference code measurements it is possible to calculate protection levels and thus thus guaranteeing the integrity of the solution [2]. Ground Tests show that the scalarized accuracy of the baseline vector is below 1m RMS We apply the double difference code technique to a system of aircraft receiver, ground maker receiver and Freewave serial data link in order to calculate the relative vector between the two GNSS Antennas. When the ground marker receiver is located on suitable landing site, this system can be used as a Beacon Landing System (BLS) where the aircraft receiver calculates deviations to a desired approach direction and approach angle. Shown on a suitable display, the pilot can follow the selected lateral and vertical track to the desired location for a successful landing or other mission. We tested this system using a setup of two Novatel OEM Star receivers running at 2Hz, one on the ground and the second one onboard the aircraft. The ground receiver is placed near the runway of our test airport in Athens, Ohio. Moreover, we provide the pilot with a display showing a horizontal situation indicator (HSI) with user-selectable approach direction and vertical angle on a small tablet computer mounted on a control yoke. The pilot selected a three degree approach angle and the runway direction of the approach end where the receiver was placed. We performed three approaches and analyzed the data using post processed dual frequency carrier phase solution with fixed integer ambiguities as a reference. The system worked well, similar to GBAS CAT1 performance without the need for precisely surveyed receiver locations but lacking the reference receiver failure redundancy. It could be used for applications, where only a relative position is relevant, such as landings of unmanned surveillance aircraft on moving platforms. Alternatively it can be applied in in situations when it is not possible to perform extensive surveying of the landing site, for example at a military forward operating bases.