Cramér-Rao Lower Bound Optimization for Hidden Moving Target Sensing via Multi-IRS-Aided Radar

Zahra Esmaeilbeig, Kumar Vijay Mishra, Arian Eamaz, Mojtaba SoltanalianIEEE Signal Processing Letters 2022 · OpenReview

TL;DR

This paper introduces a novel approach for radar systems to detect and track hidden, moving targets using multiple Intelligent Reflecting Surfaces (IRS). By optimizing the IRS phase shifts to minimize the Cramér-Rao Lower Bound (CRLB) for target parameter estimation, the proposed multi-IRS system significantly improves sensing accuracy compared to traditional radar or single-IRS setups.

What Problem Does This Paper Address?

Traditional radar systems struggle with Non-Line-of-Sight (NLoS) or "hidden" targets, where direct signal paths are blocked. While Intelligent Reflecting Surfaces (IRS) offer a promising solution by redirecting signals to overcome these obstructions, most prior work has focused on stationary targets or used only a single IRS. This paper tackles the challenge of improving the estimation accuracy of moving hidden targets using multiple IRS platforms, which is crucial for applications like autonomous driving and surveillance in complex urban environments. The core problem is how to optimally configure these multiple IRS elements to maximize the radar's ability to estimate the target's location and velocity, even when the target is not directly visible to the radar.

Methods

The authors derive the Cramér-Rao Lower Bound (CRLB) for the joint estimation of target reflectivity and Doppler velocity in a multi-IRS-aided radar system. The CRLB provides a theoretical lower limit on the achievable estimation accuracy. To optimize the IRS phase shifts, they employ an A-optimality criterion, which aims to minimize the trace of the inverse of the Fisher Information Matrix (FIM), a proxy for the CRLB. This optimization problem is non-convex, so they tackle it using an alternating optimization (AO) framework. This involves iteratively optimizing different sets of variables (IRS phase shifts and other system parameters) while keeping others fixed, which is a common strategy for solving complex optimization problems in signal processing.

Datasets

Not discussed in the paper. The evaluation is performed using simulated data based on a defined system model and parameters.

Key Insights and Ideas

  • Multiple IRS are better than one: The paper strongly suggests that deploying multiple IRS platforms offers a significant advantage over single-IRS configurations for hidden target sensing, especially when the LoS path is weak.
  • Optimizing for estimation, not just detection: Unlike some previous work that focused on maximizing signal-to-noise ratio (SNR) for detection, this paper directly optimizes IRS phase shifts to minimize the CRLB, which is directly related to estimation accuracy. This is a more targeted approach for sensing applications.
  • Doppler-aware phase shifts are key: The proposed method designs phase shifts that are aware of the target's Doppler shift, which is crucial for accurately estimating the velocity of moving targets.
  • A-optimality as a practical metric: Using the A-optimality criterion provides a tractable way to optimize the complex CRLB matrix, leading to a practical algorithm for phase shift design.

Key Results

Numerical experiments demonstrate that the proposed multi-IRS-aided radar system with optimized Doppler-aware phase shifts achieves significantly higher estimation accuracy for both target reflectivity and Doppler shifts compared to systems with no IRS or only a single IRS. The performance improvement is particularly pronounced when the LoS path loss is high (low γ). The paper shows that increasing the number of IRS platforms generally leads to better estimation accuracy.

Limitations and Future Work

  • Computational Complexity: While the alternating optimization framework helps manage complexity, the problem can still be computationally intensive, especially with a large number of IRS platforms. The paper mentions that using a generic nonlinear programming solver with multiple random initial points could increase complexity.
  • Idealized Model: The system model assumes ideal IRS elements (unimodular phase shifts) and a simplified target model (Swerling-0). Real-world IRS might have amplitude variations, and targets can exhibit more complex scattering behavior.
  • Future Work: The paper implicitly suggests extending the framework to 3D scenarios and exploring more complex target models. Further investigation into the trade-offs between the number of IRS platforms, their placement, and the resulting estimation performance would also be valuable. The authors also mention that their methods are extendable to 3D scenarios.

Summary updated

Abstract

Intelligent reflecting surface (IRS) is a rapidly emerging paradigm to enable non-line-of-sight (NLoS) wireless transmission. In this paper, we focus on IRS-aided radar estimation performance of a moving hidden or NLoS target. Unlike prior works that employ a single IRS, we investigate this problem using multiple IRS platforms and assess the estimation performance by deriving the associated Cramér-Rao lower bound (CRLB). We then design Doppler-aware IRS phase shifts by minimizing the scalar A-optimality measure of the joint parameter CRLB matrix. The resulting optimization problem is non-convex, and is thus tackled via an alternating optimization framework. Numerical results demonstrate that the deployment of multiple IRS platforms with our proposed optimized phase shifts leads to a higher estimation accuracy compared to non-IRS and single-IRS alternatives.
BibTeX
@article{esmaeilbeig2022cramer,
  author = {Zahra Esmaeilbeig and Kumar Vijay Mishra and Arian Eamaz and Mojtaba Soltanalian},
  title = {{Cramér-Rao Lower Bound Optimization for Hidden Moving Target Sensing via Multi-IRS-Aided Radar}},
  journal = {IEEE Signal Processing Letters},
  year = {2022},
  url = {https://openreview.net/forum?id=BYvoHk0NxI},
}
Figure