A Hierarchical SLAM/GPS/INS Sensor Fusion with WLFP for Flying Robo-SAR's Navigation

In this paper, we present the results of a hierarchical SLAM/GPS/INS/WLFP sensor fusion to be used in navigation system devices. Due to low quality of the inertial sensors, even a short-term GPS failure can lower the integrated navigation performance significantly. In addition, in GPS denied environments, most navigation systems need a separate assisting resource, in order to increase the availability and reliability of the device. When the GPS service/information is available, the integrated SLAM system arranges for a landmark-based map using a GPS/INS feature. But in case of inaccessibility of GPS information, the latest formerly produced map plays an important role in decreasing the INS errors. In addition, a Wireless Fingerprinting (WLFP) mechanism helps us limit the errors in the system. The results of the proposed method decreases the average estimation precision on the order of 2.6m, without any performance degradation and in different experiments, which is the maximum sustainable error (below 2.66m) for flyer landing on the base. The mentioned method could be used in computer networks to schedule the services too.