A Bayesian Surprise Approach in Designing Cognitive Radar for Autonomous Driving

Abstract

This article proposes the Bayesian surprise as the main methodology that drives the cognitive radar to estimate a target’s future state (i.e., velocity, distance) from noisy measurements and execute a decision to minimize the estimation error over time. The research aims to demonstrate whether the cognitive radar as an autonomous system can modify its internal model (i.e., waveform parameters) to gain consecutive informative measurements based on the Bayesian surprise. By assuming that the radar measurements are constructed from linear Gaussian state-space models, the paper applies Kalman filtering to perform state estimation for a simple vehicle-following scenario. According to the filter’s estimate, the sensor measures the contribution of prospective waveforms—which are available from the sensor profile library—to state estimation and selects the one that maximizes the expectation of Bayesian surprise. Numerous experiments examine the estimation performance of the proposed cognitive radar for single-target tracking in practical highway and urban driving environments. The robustness of the proposed method is compared to the state-of-the-art for various error measures. Results indicate that the Bayesian surprise outperforms its competitors with respect to the mean square relative error when one-step and multiple-step planning is considered.

Online Meta-Learning for Scene-Diverse Waveform-Agile Radar Target Tracking

“…For example, dynamic waveform selection is particularly relevant for radar scenes which are rapidly varying, heavily cluttered or include dynamic interference from neighboring systems [7]- [9]. Such scenarios are notably common in emerging vehicular radar applications [10], [11]. Additionally, if the radar application requires a high resolution, and targets can no longer be well-approximated by point scatterers, waveform diversity can provide substantial benefits in terms of target detection and classification [12]- [14].…”

An intrinsically motivated learning algorithm based on Bayesian surprise for cognitive radar in autonomous vehicles

“…In addition, Liakoni et al (2021) develops a Bayesian interpretation of surprise-based learning algorithms that modulates the rate of adaptation to new observations for estimating model parameters. In a recent paper by the same authors of this article, the Bayesian surprise is proposed as the primary approach to improving the state estimation problem in cognitive radar (Zamiri-Jafarian and Plataniotis, 2022). The research shows that minimizing the state estimation error is aligned with maximizing the expectation of Bayesian surprise, which leads to acquiring informative radar measurements.…”