Continuous Aperture Array (CAPA)-Based Secure Wireless Communications

Jingjing Zhao, Haowen Song, Xidong Mu, Kaiquan Cai, Yanbo Zhu, Yuanwei LiuIEEE Trans. Communications 2025

TL;DR

This paper explores the use of Continuous Aperture Arrays (CAPA)—a next-generation "holographic" antenna technology—to enhance wireless security. By proving that optimal transmit patterns lie within a tiny "channel subspace," the authors transform an intractable infinite-dimensional math problem into an efficient algorithm that significantly outperforms traditional discrete MIMO systems in both secrecy rate and computational speed.

What Problem Does This Paper Address?

Traditional Multiple-Input Multiple-Output (MIMO) systems rely on discrete antenna elements spaced apart (usually by half a wavelength). This physical spacing limits the "spatial degrees of freedom"—essentially the number of independent paths a signal can take.

Continuous Aperture Arrays (CAPA) are the ultimate evolution of this, providing a nearly continuous electromagnetic surface where the amplitude and phase of the current can be controlled at every point. While CAPA offers massive theoretical gains, it introduces a mathematical nightmare: instead of optimizing a finite vector of antenna weights, you have to optimize a continuous current pattern function. This is a "functional programming" problem with an infinite number of variables.

The authors specifically tackle Physical Layer Security (PLS). They want to maximize the "secrecy rate" (the speed of secure data transmission) to a legitimate user while simultaneously jamming an eavesdropper with "Artificial Noise" (AN), all while staying within a power budget.

Methods

The researchers propose two primary ways to solve the CAPA beamforming problem:

  1. Channel Subspace-Based Approach: This is the paper's core innovation. The authors mathematically prove that the optimal current pattern for the signal and the noise must lie within the subspace spanned by the channel responses of the legitimate user and the eavesdropper. This "dimensionality reduction" allows them to convert an infinite-dimensional integral problem into a simple optimization of a few weighting factors. They solve this using a Penalty-based Successive Convex Approximation (SCA) method.
  2. Two-Stage ZF-MRT Scheme: For even lower complexity, they propose a heuristic. In stage one, they use Zero-Forcing (ZF) to ensure the jamming noise doesn't interfere with the legitimate user and Maximum Ratio Transmission (MRT) to beam the signal directly to the user. In stage two, they use a simple one-dimensional search to find the best power split between the signal and the noise.
  3. Baseline Comparison: They compare their approach against the state-of-the-art Fourier-based approach, which approximates the continuous current using a finite sum of Fourier series.

Datasets

As a theoretical communications paper, no external datasets were used. Instead, the authors conducted extensive numerical simulations using:

  • Electromagnetic Channel Models: Based on Green’s functions for a homogeneous medium.
  • Synthetic Scenarios: Randomly placed users and eavesdroppers within a defined 3D volume.
  • Parameters: Simulations were run at various frequencies (2.4 GHz to 15 GHz) and aperture sizes (0.25 $m^2$ to 0.5 $m^2$).

Key Insights and Ideas

  • The Power of Subspace: The most significant "aha" moment is the proof that you don't need infinite Fourier terms to describe the optimal CAPA pattern. Because the "physics" of the channel is defined by the locations of the users, the optimal solution is naturally constrained to those specific spatial directions.
  • Complexity Independence: Unlike Fourier methods, where the number of variables explodes as the frequency or aperture size increases, the subspace method's complexity remains constant. It only depends on the number of users, not the size of the antenna.
  • CAPA vs. Discrete MIMO: CAPA isn't just "MIMO with more antennas." By filling the gaps between discrete elements, CAPA captures spatial nuances that discrete arrays miss, leading to much higher secrecy rates, especially when the eavesdropper is physically close to the legitimate user.

Key Results

  • Secrecy Gains: CAPA achieved up to an 87% improvement in secrecy rate over conventional discrete MIMO systems.
  • Computational Efficiency: The proposed subspace algorithm is orders of magnitude faster than Fourier methods. For a 0.5 $m^2$ aperture at 7.8 GHz, the subspace method took ~1.9 seconds, while the Fourier method took over 532 seconds (nearly 280x slower).
  • Heuristic Performance: The low-complexity ZF-MRT approach was found to be nearly as good as the optimal subspace method in high-power scenarios, making it a viable candidate for real-time hardware implementation.

Limitations and Future Work

  • Perfect CSI Assumption: The paper assumes the base station has perfect Channel State Information (CSI) for both the user and the eavesdropper. In reality, eavesdroppers are usually "passive" and don't help you estimate their channel.
  • Single-User Focus: The model currently handles one legitimate user and one eavesdropper. Extending this to multi-user "broadcast" scenarios with multiple colluding eavesdroppers is a necessary next step.
  • Hardware Realization: While the math works, building a surface that allows truly continuous current control is a massive materials science and circuit design challenge (e.g., using dynamic metasurfaces). The paper doesn't account for specific hardware impairments like phase noise or mutual coupling in extremely dense circuits.

Summary updated

Abstract

A continuous aperture array (CAPA)-based secure communication system is investigated, where a base station equipped with a CAPA transmits signals to a legitimate user under the existence of an eavesdropper. For improving the secrecy performance, the artificial noise (AN) is employed at the BS for the jamming purpose. We aim at maximizing the secrecy rate by jointly optimizing the information-bearing and AN source current patterns, subject to the maximum transmit power constraint. To solve the resultant non-convex integral-based functional programming problem, a channel subspace-based approach is first proposed via exploiting the result that the optimal current patterns always lie within the subspace spanned by all users' channel responses. Then, the intractable CAPA continuous source current pattern design problem with an infinite number of optimization variables is equivalently transformed into the channel-subspace weighting factor optimization problem with a finite number of optimization variables. A penalty-based successive convex approximation method is developed for iteratively optimizing the finite-size weighting vectors. To further reduce the computational complexity, we propose a two-stage source current patterns design scheme. Specifically, the information-bearing and AN patterns are first designed using the maximal ration transmission and zero-forcing transmission, respectively. Then, the remaining power allocation is addressed via the one-dimensional search method. Numerical results unveil that 1) the CAPA brings in significant secrecy rate gain compared to the conventional discrete multiple-input multiple-output; 2) the proposed channel subspace-based algorithm outperforms the conventional Fourier-based approach, while sustaining much lower computational complexity; and 3) the two-stage ZF-MRT approach has negligible performance loss for the large transmit power regime.
BibTeX
@article{zhao2025continuous,
  author = {Jingjing Zhao and Haowen Song and Xidong Mu and Kaiquan Cai and Yanbo Zhu and Yuanwei Liu},
  title = {{Continuous Aperture Array (CAPA)-Based Secure Wireless Communications}},
  journal = {IEEE Trans. Communications},
  year = {2025},
}
Figure