In the face of emerging Global Navigation Satellite System (GNSS) spoofing attacks, there is a need to give a comprehensive analysis on how the inertial navigation system (INS)/GNSS integrated navigation system responds to different kinds of spoofing attacks. A better understanding of the integrated navigation system’s behavior with spoofed GNSS measurements gives us valuable clues to develop effective spoofing defenses. This paper focuses on an impact assessment of GNSS spoofing attacks on the integrated navigation system Kalman filter’s error covariance, innovation sequence and inertial sensor bias estimation. A simple and straightforward measurement-level trajectory spoofing simulation framework is presented, serving as the basis for an impact assessment of both unsynchronized and synchronized spoofing attacks. Recommendations are given for spoofing detection and mitigation based on our findings in the impact assessment process.

Since the conventional strapdown inertial navigation system (SINS) and odometer (OD) integrated navigation has certain limitations, such as suffering from scale factor variation and being affected by wheel slipping and skidding, we present a high-accuracy integrated navigation system that uses both the SINS and laser Doppler velocimeter (LDV) for land vehicles to overcome these problems. We took the lever-arm residual and boresight errors into consideration and formed a complete 21-state Kalman filter to accomplish integrated navigation and precise sensor-to-sensor calibration. The observability analysis of the SINS/LDV system provides valuable insights into the conditions under which the states can be estimated. In addition, the summed measurement model and sequential Kalman filter calculation are deduced to make full use of the non-holonomic constraints (NHC) and LDV measurements. Finally, a long-distance experiment (230 km) with a navigation-grade IMU demonstrated that positioning accuracies of better than 0.02% of the distance travelled can be achieved in both the horizontal and vertical directions.

This paper describes a fully autonomous real-time in-motion alignment algorithm for Strapdown Inertial Navigation Systems (SINS) in land vehicle applications. Once the initial position is available, the vehicle can start a mission immediately with accurate attitude, position and velocity information determined within ten minutes. This is achieved by two tightly coupled stages, that is, real-time Double-vector Attitude Determination Coarse Alignment (DADCA) and Backtracking Fine Alignment (BFA). In the DADCA process, the vehicle motion is omitted to roughly estimate the attitude at the very start of the alignment. Meanwhile, attitude quaternions and velocity increments are extracted and recorded. The BFA process utilises the stored data and exploits the Non-Holonomic Constraints (NHC) of a vehicle to obtain virtual velocity measurements. A linear SINS/NHC Kalman filter with mounting angles as extended states is constructed to improve the fine alignment accuracy. The method is verified by three vehicle tests, which shows that the accuracy of alignment azimuth is 0·0358° (Root Mean Square, RMS) and the positioning accuracy is about 15 m (RMS) at the end of the alignment.

For situations where accurate attitude information is needed while position is not readily available or not necessarily required, such as the attitude heading reference system (AHRS), this paper proposes a velocity-aided in-motion alignment scheme for a strapdown inertial navigation system (SINS) during geographic latitude uncertainty. In the coarse alignment stage, a gravitational apparent motion rebuilding method based on integral velocity smoothing (GAM-IVS) is designed to roughly determine the attitude and latitude information. Then, the SINS/velocity-integrated navigation method uses the recorded data to complete the fine alignment. Due to the adoption of a well-designed backtracking navigation strategy, the storage space and computation load are reduced to a degree that an ordinary navigation computer can work with. Nine ship tests were carried out and the results show that the proposed method can align the azimuth with an accuracy of 0.04° in a moving environment, which is the same as that of normal in-motion alignment with known position.

This paper presents a reliable in-motion alignment algorithm for a low cost Strapdown Inertial Navigation System/Global Positioning System (SINS/GPS) combination under random misalignment angles, which transforms attitude alignment into an attitude estimation problem. Based on Rodrigues parameters, an alignment model with a linear state-space equation and a second order nonlinear measurement equation is established. Furthermore, by employing a Taylor expansion on the nonlinear measurement equation, we implement a second order Extended Kalman Filter (EKF2). The proposed method uses a single filter that can not only determine the initial attitude, but also estimate the sensor errors. In addition, a scheme is given for avoiding singularity, which makes the algorithm more widely suitable for random misalignment angles. Experimental ground tests are performed with a low-cost Micro-Electromechanical System (MEMS) SINS, which validates the efficacy of the proposed method. The performance compared to the traditional alignment algorithm is also given.