Inter-Sensor Validation for Improved Reliability of Multi-Sensor Systems

Modern day personal navigation devices are often realized as a multi-sensor integrated navigation system wherein the measurements from various sensors are combined to generate a navigation solution. A fault occurring in any of these sensors can affect the performance of the navigation system, potentially compromising the reliability of the navigation solution. It is thus imperative to detect and remove faults occurring in the individual sensors before they are integrated in a navigation filter. In an attempt to improve the reliability of the navigation solution, this paper proposes a sensor validation scheme that exploits the analytical redundancy between the dissimilar sensors used in multi-sensor integrated systems to detect and remove the faults in individual sensor measurements. The paper identifies possible fault scenarios in the sensors being used and characterizes some of the sensor faults through experimental analysis. The proposed scheme is then analyzed using both simulated and real sensor fault scenarios. In both scenarios, the proposed scheme successfully detects the sensor faults. Moreover, the subsequent implication of the fault detection scheme in the position solution is also analyzed.