Adversarial Radar Inference: Inverse Tracking, Identifying Cognition, and Designing Smart Interference

Abstract

This paper considers three inter-related adversarial inference problems involving cognitive radars. We first discuss inverse tracking of the radar to estimate the adversary's estimate of us based on the radar's actions and calibrate the radar's sensing accuracy. Second, using revealed preference from microeconomics, we formulate a non-parametric test to identify if the cognitive radar is a constrained utility maximizer with signal processing constraints. We consider two radar functionalities, namely, beam allocation and waveform design, with respect to which the cognitive radar is assumed to maximize its utility and construct a set-valued estimator for the radar's utility function. Finally, we discuss how to engineer interference at the physical layer level to confuse the radar which forces it to change its transmit waveform. The levels of abstraction range from smart interference design based on Wiener filters (at the pulse/waveform level), inverse Kalman filters at the tracking level and revealed preferences for identifying utility maximization at the systems level.

Meta-Cognition. An Inverse-Inverse Reinforcement Learning Approach for Cognitive Radars

“…When state x n represents the position and velocity in Euclidean space, A is a block diagonal constant velocity matrix [33]. The state noise covariance Q(α) in (10) models acceleration maneuvers of the target parameterized by the probes α.…”

“…Assuming the model parameters (10) satisfy the conditions that [A, C] is detectable and [A, √ Q] is stabilizable, the steadystate predicted covariance Σ ∞ is the unique positive semidefinite solution of the algebraic Riccati equation (ARE):…”

“…A cognitive radar [6], [7], [8] uses the perception-action cycle of cognition to sense the environment and learn from it relevant information about the target and the environment. The cognitive radar is a reinforcement learner that maximizes its utility [9], [10] and tunes its sensor to optimally satisfy its mission objectives. In the context of DFIG (Data Fusion Information Group) process model [11], sensor adaptation by the radar can be viewed as Level 4-Process Refinement in the DFIG model.…”

Inverse-Inverse Reinforcement Learning. How to Hide Strategy from an Adversarial Inverse Reinforcement Learner

“…Theorem 3 specified the procedure for a decision maker to effectively mask its cognition from an adversary. Here, we apply our I-IRL result to the problem of a cognitive radar optimizing waveform based on the SINR of the adversarial target measurement [23]. The adversary observes the radar over k = 1, 2, .…”

“…Clearly, the above setup falls under the non-linear utility maximization setup in Definition 1. For appropriately chosen matrices (see [23]), the utility in (19) can be shown to be monotonically increasing in β.…”