[Paper Review] Estimation of Angles of Arrival Through Superresolution -- A Soft Recovery Approach for General Antenna Geometries

Mahdi Barzegar, Guiseppe Caire+4 more|arXiv (Cornell University)|Nov 10, 2017

Sparse and Compressive Sensing Techniques20 references3 citations

TL;DR

This paper proposes a superresolution-based direction-of-arrival (DoA) estimation method using total variation (TV) minimization for general 2D antenna array geometries, extending prior work limited to uniform linear arrays. It establishes theoretical recovery guarantees via a soft recovery framework, showing that DoAs can be approximately recovered when angular separation and signal power conditions are met, with numerical validation across diverse array designs.

ABSTRACT

The estimation of direction of arrivals with help of $TV$-minimization is studied. Contrary to prior work in this direction, which has only considered certain antenna placement designs, we consider general antenna geometries. Applying the soft-recovery framework, we are able to derive a theoretic guarantee for a certain direction of arrival to be approximately recovered. We discuss the impact of the recovery guarantee for a few concrete antenna designs. Additionally, numerical simulations supporting the findings of the theoretical part are performed.

Motivation & Objective

• To extend superresolution DoA estimation beyond uniform linear arrays to general 2D antenna geometries.
• To provide theoretical recovery guarantees for DoA estimation using TV-minimization under continuous DoA models.
• To apply the soft recovery framework to derive conditions under which a true DoA is approximately recovered.
• To analyze the impact of array geometry on recovery performance through theoretical and numerical studies.
• To validate the theoretical findings with numerical simulations across multiple antenna designs.

Proposed method

• Formulates DoA estimation as a continuous measure recovery problem using a discrete complex measure $\mu_0 = \sum_{\ell=1}^s w_\ell \delta_{u_\ell}$ over the unit circle.
• Uses TV-norm minimization as a convex relaxation to promote sparsity in the measure, solving $\min \|\mu\|_{TV} \text{ s.t. } \widehat{\mu}_k = f_k$.
• Applies the soft recovery framework from [11] to derive sufficient conditions for approximate recovery of a DoA peak.
• Introduces a parameter $\gamma(R)$ dependent on array geometry and signal power to quantify recovery feasibility.
• Derives a condition $\gamma(R) + R^{-k} \leq C|c_{\theta_0}|$ for successful recovery, where $C$ is a universal constant.
• Validates results via numerical simulations across various 2D antenna geometries, including non-uniform linear and circular arrays.

Experimental results

Research questions

• RQ1 Can TV-minimization reliably recover DoAs in general 2D antenna array geometries beyond uniform linear arrays?
• RQ2 What theoretical conditions ensure that a true DoA is approximately recovered by the TV-minimization framework?
• RQ3 How does the array geometry influence the recovery performance and the required signal-to-noise ratio or angular separation?
• RQ4 What is the role of the soft recovery framework in establishing recovery guarantees for continuous DoA estimation?
• RQ5 How do different antenna designs (e.g., non-uniform linear, circular) affect the recovery parameters and performance?

Key findings

• Theoretical recovery is guaranteed when the angular separation between DoAs satisfies a minimum distance condition and the signal power at the target DoA exceeds a threshold determined by array geometry and $\gamma(R)$.
• The recovery condition $\gamma(R) + R^{-k} \leq C|c_{\theta_0}|$ ensures that any minimizer of the TV program has a peak near the true DoA $\theta_0$.
• For non-uniform linear arrays with $m \geq 5$ elements, the recovery condition holds when $\theta_0 < 0.6378$ radians, corresponding to $m \geq 5$.
• Numerical simulations confirm the theoretical findings, showing stable recovery across diverse 2D array geometries including circular and rectangular arrays.
• The soft recovery framework successfully extends recovery guarantees to arbitrary antenna geometries, overcoming limitations of prior grid-based or uniform array-only approaches.
• The recovery quality depends on the relative power $|c_{\theta_0}|$, the choice of $R$, and the geometry-dependent parameter $\gamma(R)$.