Aiming at the background of poor positioning effect and high positioning demand in GNSS rejection environments such as underground parking lots and mines, a BLE/INS/ODO tight-coupled navigation algorithm based on signal attenuation model is proposed. In the algorithm, the autonomous integrity monitoring of RSSI signals in motion is firstly completed, and the observed abnormal data are eliminated based on the residual chi-square test and random search method; then the tight combined extended Kalman filter of 22 states of BLE/INS/ODO positioning system is established considering the dynamic environment fading factor error; the overall Sage-Husa anti-difference adaptive filtering is applied to complete the measurement noise variance array adaptive. Meanwhile, the experimental results show that the algorithm can effectively compensate for the shortcomings of INS/ODO autonomous positioning, while suppressing the effects of Bluetooth signal noise caused by signal occlusion and multi-path effects. Compared with the combination of RSSI great likelihood positioning, INS/ODO combination and traditional Bluetooth positioning, the navigation accuracy of the system obtained by this algorithm is greatly improved.