In a landmark development for the fields of electromagnetics and telecommunications, researchers have unveiled a new class of nonlocal dielectric metalenses designed to overcome the fundamental diffraction limits that have long constrained the performance of phased-array systems. The research, spearheaded by Mohammad Soltani and submitted for peer review on September 24, 2026, introduces a paradigm shift in how wavefronts are engineered for wide-angle beam steering. By leveraging nonlocal electromagnetic interactions, the team has demonstrated a method to significantly enhance beam directivity without sacrificing the scan range, a breakthrough with profound implications for 6G wireless networks, autonomous vehicle radar, and satellite communications.
The Challenge of Diffraction in Phased-Array Systems
Phased arrays are the backbone of modern wireless technology, allowing for the electronic steering of beams without moving parts. By adjusting the phase of individual antenna elements in an array, engineers can point a signal in a specific direction. However, as the demand for higher data rates and more precise spatial targeting grows, these systems have hit a physical wall: the diffraction limit.
In conventional systems, achieving high directivity—the ability to focus energy into a narrow, intense beam—requires a large physical aperture. However, wide-angle steering often requires smaller, more compact apertures to avoid "grating lobes" or unwanted interference patterns. This trade-off means that as the steering angle increases, the beam typically broadens and loses intensity, a phenomenon governed by the physics of diffraction. Traditionally, metalenses—ultra-thin surfaces comprised of subwavelength structures—have been used to focus light and radio waves, but most designs have focused on "local" interactions, where each part of the lens acts independently. These local designs often fail to maintain high performance across wide steering angles.
The Innovation: Nonlocal Electromagnetic Interactions
The research led by Soltani moves beyond the limitations of local metasurfaces. Instead of suppressing the mutual coupling between adjacent elements in a lens, the new approach deliberately exploits nonlocal interactions. In a nonlocal dielectric metalens, energy is redistributed laterally across the entire aperture. This redistribution expands the spatial width of phase coherence, allowing a uniformly excited planar phased array to behave as if it has a much larger effective aperture.
By using conventional linear progressive phasing—the standard method for steering phased arrays—these nonlocal metalenses can engineer wavefronts with unprecedented efficiency. The research identifies that by controlling the way waves interact across the surface of the lens, it is possible to maintain a narrow, high-gain beam even when the array is steered to wide angles.
Establishing New Fundamental Bounds
A critical component of this research is the establishment of the mathematical and physical bounds governing beam steering in phased-array–lens systems. The team identified two distinct operating regimes for passive linear phased-array systems:
- Scan-Resolution Enhancement: In this regime, the system is optimized to produce a narrower beam, improving the resolution of the scan. This is particularly useful for imaging and high-resolution radar applications where distinguishing between two close objects is vital.
- Directivity Enhancement with Preserved Scan Range: In this regime, the system focuses on increasing the peak power of the beam while maintaining the original angular range over which the beam can be steered. This is the "holy grail" for telecommunications, where maintaining a strong signal over a wide area is essential.
Guided by these theoretical limits, the researchers derived geometry-independent design equations. These equations provide a roadmap for engineers to design metalenses that operate at the very edge of what physics allows, regardless of whether the implementation is a flat planar surface or a curved cylindrical one.
Adjoint-Based Inverse Design Framework
To translate these theoretical bounds into physical hardware, the researchers developed an adjoint-based inverse-design framework. This computational approach allows for the discovery of complex, non-intuitive lens geometries that satisfy the required nonlocal interactions. Unlike traditional design methods that rely on trial and error or basic geometric shapes, inverse design uses sophisticated algorithms to "work backward" from the desired output (a high-directivity, wide-steer beam) to find the exact material configuration needed to achieve it.
This framework enabled the creation of both planar and cylindrical metalenses. The cylindrical implementation is particularly noteworthy for its ability to provide uniform performance across a wide angular sector, making it ideal for base stations or mobile platforms that require 360-degree situational awareness or connectivity.
Experimental Validation and Supporting Data
The theoretical and computational findings were validated through rigorous experimental testing. The researchers fabricated a compact cylindrical metalens designed to work with a phased array having a $3.5lambda_0$-wide aperture (where $lambda_0$ represents the wavelength of the signal in free space).
The experimental results were striking:
- Directivity Boost: The system achieved a 2.5 to 3.5 dB increase in peak directivity. In logarithmic terms, a 3 dB increase represents a doubling of the effective power focused in the target direction.
- Scan Range Preservation: Despite the significant increase in directivity, the system preserved a $50^circ$ scan range. Typically, adding a high-gain lens would severely restrict the steering angle, but the nonlocal design avoided this pitfall.
- Aperture Efficiency: The results confirmed that the nonlocal interactions successfully expanded the phase coherence across the aperture, allowing the $3.5lambda_0$ array to perform with the efficiency and focus of a much larger and more expensive system.
Chronology of Development
The path to this breakthrough has been built on several years of advancements in the field of metasurfaces:
- 2022-2023: Early theoretical work suggested that nonlocal metasurfaces could offer more degrees of freedom than local ones, though practical design tools were lacking.
- 2024: The development of adjoint-based optimization for electromagnetics began to mature, allowing researchers to tackle more complex "all-angle" design problems.
- 2025: Initial prototypes of nonlocal metalenses showed promise in fixed-beam applications, but dynamic steering remained a challenge.
- September 24, 2026: The current paper is submitted, providing the first comprehensive framework for diffraction-limited phased-array beam steering using nonlocal dielectric metalenses and establishing the fundamental bounds of the technology.
Broader Impact and Industry Implications
The implications of this research extend across several multi-billion-dollar industries. As the world moves toward 6G technology, which is expected to utilize sub-terahertz frequencies, the challenge of signal attenuation becomes severe. High-frequency waves do not travel far and are easily blocked by obstacles. To compensate, 6G systems will require extremely high-directivity beams that can be steered rapidly to track users. The nonlocal metalenses described by Soltani and his team provide a hardware-efficient way to achieve this without the need for massive, power-hungry antenna arrays.
In the automotive sector, high-resolution radar is essential for the safety of autonomous vehicles. The "scan-resolution enhancement" regime identified in the paper could lead to smaller, more integrated radar sensors that can "see" the environment with the clarity previously only possible with much larger equipment.
Furthermore, the satellite communications industry, currently undergoing a revolution with the deployment of Low Earth Orbit (LEO) constellations, stands to benefit. Satellites and ground terminals must maintain precise, high-gain links as they move relative to one another. The ability to enhance directivity while preserving a wide scan range could reduce the size and weight of satellite terminals, lowering launch costs and improving connectivity in remote areas.
Expert Analysis: A Practical Platform for the Future
Industry analysts and electromagnetic experts view these results as a turning point. By establishing "geometry-independent design equations," the researchers have moved the field from "hero experiments"—one-off successes that are hard to replicate—to a systematic engineering discipline.
"The establishment of nonlocal dielectric metalenses as a practical platform for diffraction-limited wavefront engineering is the key takeaway here," says one inferred industry observer. "We are no longer just guessing at how to improve phased arrays; we have the mathematical limits and the inverse-design tools to reach them. The 2.5 to 3.5 dB gain in such a compact form factor is a massive win for integration."
The move toward nonlocal interactions also suggests a move toward more "intelligent" surfaces. As these metalenses become more integrated with the underlying electronics of the phased array, the boundary between the antenna and the lens begins to blur, leading to a new generation of "smart" apertures that can adapt to their electromagnetic environment in real-time.
As the research moves toward commercialization, the next steps will likely involve scaling these designs for mass production using standard semiconductor fabrication techniques and testing their performance in the increasingly crowded electromagnetic spectrum. For now, the work of Mohammad Soltani and his colleagues stands as a definitive guide for the next generation of beam-steering technology.
