핵심 개념
This paper develops efficient inverse cubature Kalman filter (I-CKF), inverse quadrature Kalman filter (I-QKF), and inverse cubature-quadrature Kalman filter (I-CQKF) to estimate the state inferred by an adversarial cognitive radar. The proposed methods can handle highly non-linear system models where extended Kalman filter's linearization often fails. The paper also derives stability and consistency conditions for the inverse filters and demonstrates their improved estimation accuracy compared to the recursive Cramér-Rao lower bound.
초록
This paper addresses the inverse cognition problem, where a 'defender' (e.g., an intelligent target) aims to estimate the state inferred by an 'attacker' (e.g., a cognitive radar) using noisy observations of the attacker's actions. The authors develop efficient inverse filters based on cubature, quadrature, and cubature-quadrature Kalman filtering techniques to handle highly non-linear system models.
The key highlights are:
- Inverse CKF (I-CKF), I-QKF, and I-CQKF are proposed to estimate the attacker's state inference, which outperform the inverse extended Kalman filter (I-EKF) in case of severe non-linearities.
- Stability and consistency conditions are derived for the proposed inverse filters, showing that the forward filter's stability is sufficient to guarantee the same for the inverse filter under mild conditions.
- For unknown system models, a reproducing kernel Hilbert space (RKHS)-based CKF is developed to jointly estimate the state and learn the unknown system parameters.
- Numerical experiments demonstrate the improved estimation accuracy of the proposed inverse filters compared to the recursive Cramér-Rao lower bound.
- The inverse filters are also extended to handle non-Gaussian noise, continuous-time state evolution, and complex-valued systems.
The paper provides a comprehensive framework for efficient inverse Bayesian filtering in cognitive radar systems with highly non-linear dynamics.
통계
The paper does not provide any specific numerical data or metrics. However, it presents the theoretical derivations and performance analyses of the proposed inverse filters.
인용구
"Recent research in inverse cognition with cognitive radar has led to the development of inverse stochastic filters that are employed by the target to infer the information the cognitive radar may have learned."
"In this paper, we consider the efficient numerical integration techniques to address such non-linearities and, to this end, develop inverse cubature KF (I-CKF), inverse quadrature KF (I-QKF), and inverse cubature-quadrature KF (I-CQKF)."
"Our theoretical analyses show that the forward filter's stability is sufficient to guarantee the same for the inverse filter under mild conditions imposed on the system."