An Ultra-tightly coupled GPS/INS Integration using Federated Kalman Filter

Traditional GPS/INS integration designs use a loosely or a tightly coupled architecture in which the GPS receiver channels independently track the broadcast satellite signal and constructs the estimated range and range rate to the satellite. A GPS/INS navigation filter blends the GPS pseudorange and inertial measurements and obtains the vehicle navigation solution. In tightly coupled system, the pseudorange rate information is added into the tracking loop of the receiver. This architecture is sub-optimal from the standpoint of preventing signal loss due to jamming and high vehicle dynamics. A deeper level of integration technique is called ultra tightly coupling. The ultra-tightly coupled architecture combines the receiver signal tracking loops and the IMU and the navigation filter function into a single integrated filter. In this case, the filter operates on the receiver-tracking loop I and Q signals which are at a basic level than the pseudorange and delta range and the IMU measurements to estimate navigation information. The federated filter comprises an extension of Kalman filtering to distribute estimation problems characterized by a collection of local estimators whose solutions are combined by a global estimator. The advantage of this filter is the reduced states of filter and the reduced calculation load caused by the separation of local and master filter. Additionally, this filter can detect easily the fault of sensor independently. In this paper, we propose a new method for ultra-tightly coupled GPS/INS integration using federated Kalman filter. In this architecture, the GPS/INS navigation filter represents the federated global filter and the signal tracking filters represent the local filters. The local filters directly process the GPS receiver channel I and Q measurements and IMU measurements to estimate the signal power, pseudorange, Doppler, and vehicle position. The estimated local state variable and its covariance are fed into the master GPS/INS navigation filter and global optimal data fusion is performed.