Inverse Cognition in Nonlinear Sensing Systems
Himali Singh, Arpan Chattopadhyay, Kumar Vijay Mishra · 2022
Inverse cognition aims to estimate the ‘estimate’ of a defender's state inferred by an adversarial cognitive system such as radar when the defender observes the attacker's actions in a noisy environment. We develop an inverse extended Kalman filter (IEKF) to address inverse cognition in a non-linear system setting. We consider general non-linear system dynamics, wherein the attacker employs an extended Kalman filter to compute its estimate of the defender's state. We further provide sufficient conditions for the boundedness of the IEKF's estimation error. We validate our model and methods through numerical experiments using recursive Cramer- Rao lower bound as a benchmark.