For outage of GPS, a multi-radio integrated navigation system has been considered, in which Global Navigation Satellite System (GNSS) and ground-based radio navigation systems are integrated. By using Kalman filter, the multi-radio integrated navigation system can provide accurate integrated navigation output even when the vehicle is in high dynamic motion. Since navigation system receives lots of measurements from various navigation signal sources, the measurement matrix of the Kalman filter becomes large. As a result of this, larger computational capability can be required. And a fault of a subsystem in the multi-radio integrated navigation system may affect overall operation of the integrated navigation system. It is known that computational load can be reduced and the influence of the fault of subsystems can be blocked when the federated Kalman filter and/or the decentralized Kalman filter are used (Carlson 1988, Carlson & Berarducci 1994). In this paper, a design method of a federated Kalman filter for the multi-radio integrated navigation system is proposed. Using the proposed method, a federated Kalman filter is implemented in a multi-thread configuration for the M&S software of the multi-radio integrated navigation system. For performance evaluation of the proposed method, navigation error is compared with those of the centralized Kalman filter.

Keywords: Federated filter, Kalman filter, integrated navigation system