spoofing attack detection in integrated gnss/ins navigation system using self-tuning kalman filter

Spoofers may attack global navigation satellite system (gnss) data, resulting in positioning errors in navigation. the features of the prediction of the kalman filter are utilized to detect spoofing attacks in most of the existing attack monitors. however, the basic assumption in the conventional kalman filter is that the noise characteristics in the system model are known in advance, which is often violated in real-world navigation systems. so, a novel spoofing attack monitor for a loosely coupled integrated gnss/ins system is developed here utilizing the self-tuning kalman filter concept in which the covariance matrices of the dynamic and measurement noises are adapted online, during the operation of the system. the proposed attack detector is designed based on the covariance matching idea in the adaptive kalman filter; wherein, the covariance matrices of the noise models are adapted based on the residual and innovation sequences within a moving sampling window. comparative simulations demonstrate that the suggested method significantly outperforms the one based on the conventional kalman filter.