The case of a land vehicle traveling in a straight and leveled trajectory with a position aided inertial navigation system (INS) is considered. In such a scenario, the heading angle and some of the inertial sensors error states are not observable. To circumvent this problem, we propose a method to extract attitude information from position measurements history provided by the external sensor. Simultaneously utilizing the attitude information and position updates in the navigation filter enables the estimation of the heading angle and all of the inertial sensors error states. That is, using the proposed approach the system becomes completely observable. Simulation and field experiment results are provided to show the navigation solution performance when applying the proposed approach.

Due to the need of providing navigation information around the clock independently, the Inertial Navigation System (INS)/Synthetic Aperture Radar (SAR) integrated navigation system is a hot topic currently. This paper proposes a deep fusion method based on INS/SAR integrated navigation. Different from the conventional INS/SAR integrated navigation system, the state equation and the measurement equation are established, regarding the SAR pseudo-range and pseudo-range rate bias error as independent state variables. A result of qualitative analysis is provided, by comparing the one side of distribution and the two sides of distribution of measurement points. The simulation result shows that the position error of two sides of distribution greatly reduces. Compared with the one side of distribution, the accuracy improves approximately 78%. The pseudo-range and pseudo-range rate bias error can be estimated effectively with two sides of distribution, of which the accuracy is 69% and 76% approximately, improving 39% and 64% respectively compared with the one side of distribution. The result of flight experiments verifies the effectiveness of the proposed method. The position accuracy of measurement points based on two sides of the distribution is better than that of the distribution of one side, which improves the actual accuracy of SAR, and is consistent with the simulation results.