In the Inertial Navigation System/Geomagnetic Navigation System (INS/GNS) integrated navigation system, divergence and measurement noise that change with environments frequently occur must be reduced so as to extend the accuracy and stability of autonomous, passive navigation. Base on the Unscented Kalman Filter, we can combine Convergence of intelligent sensing as well as computing, and control for UAV nabled B5G/6G network to deal with change in the variance and mean value of the latest information. First，the noise’s covariance in the model is modified “online” to change the estimation of mean square deviation error and filtering gain of Kalman filtering by using 5G/6G network; Then, through UAV enabled B5G/6G network and intelligent sensing to change the scaling factor of weight in sigma sampling is adaptively changed to solve the nonlocal effect in UT transform and improve the effciency and accuracy of the navigation deposition system. The simulation findings reveal that the fuzzy adaptive Kalman filter is very efficient for INS/ GNS integrated navigation systems. I tover come the shortcomings of traditional filtering method and improve the accuracy of filtering.