For the first time, researchers have developed a technology that could enable 6G antennas to be mass-produced on wafers in much the same way as semiconductor devices.
A research team led by Professor Jungsuek Oh of the Department of Electrical and Computer Engineering at Seoul National University College of Engineering has developed the world’s first wafer-integrated 6G beamforming antenna technology, dubbed STARE, in which antennas and RF semiconductor switches that control the direction of radio waves are fabricated simultaneously on a single wafer.
Conventionally, antennas and the RF semiconductor chips that control electromagnetic waves have had to be fabricated separately and then assembled. The SNU team, however, succeeded in integrating both functions within a single wafer. Much like combining a washing machine and dryer into a single appliance, the new approach makes it possible to manufacture two previously separate components through a single semiconductor fabrication process. The technology lays the groundwork for producing the ultra-high-density antenna arrays required for 6G and satellite communications in smaller, more cost-effective form factors.
The findings were published in the internationally renowned journal Nature Communications .
6G mobile communications and satellite communications require far more antennas than existing systems, with multiple antennas simultaneously directing electromagnetic waves toward desired locations. To meet this need, reconfigurable intelligent surfaces (RIS), which act like “smart mirrors” that reflect or refract electromagnetic waves in desired directions, as well as new types of active antenna technologies, have recently attracted considerable attention.
Conventional RIS technologies, however, require antennas and electromagnetic-wave control components to be fabricated separately and then assembled, making manufacturing complex and costly. Liquid crystals have slow response speeds and are difficult to fabricate, while PIN diodes and varactors suffer from high-frequency losses and add surface-mounted component volume. As the number of antennas increases, so do the number of components, interconnections, and assembly steps, creating a major obstacle to the implementation of large-scale 6G systems.
Semiconductors and antennas also differ substantially in both size and design methodology, making them difficult to fabricate through a single process. Semiconductor devices are generally designed on the nanometer scale, whereas 6G antennas employ millimeter-scale structures determined by wavelength. In addition, semiconductor design software and electromagnetic simulation software for antennas are fundamentally different, and there has been no design tool capable of handling both domains simultaneously. Researchers also had to overcome the high dielectric constant and radio-frequency losses of silicon, as well as the need for costly processes that create connections through the wafer.
To overcome these limitations, the research team proposed a new architecture called STARE (Semiconductor Transmit Array with Reconfigurable Elements). Whereas conventional RIS can be viewed as an “assembled antenna device,” in which electromagnetic-wave control components are attached one by one to completed antennas, STARE is a “semiconductor-integrated antenna device” in which antennas and RF semiconductor switches are fabricated simultaneously on the same wafer. Because the number of discrete components does not have to increase dramatically as the antenna array grows, the architecture is well suited to mass production.
To realize the architecture, the team optimized the structure of the RF semiconductor switch and established a co-design methodology that incorporates actual measured device characteristics into electromagnetic design. This enabled the semiconductor components and antennas to be designed as a single integrated system.
The researchers also developed a new architecture that uses magnetic coupling between metal structures on opposite sides of the wafer, eliminating the need for costly through-wafer interconnection processes. This reduced energy losses and enabled a planar beam-steering device in which the integrated semiconductor structures do not protrude from the surface. In the final prototype, the researchers successfully demonstrated the entire process—from design and fabrication to electronic control and actual beam steering.
The significance of the research lies not simply in the creation of a new antenna, but in the introduction of a new manufacturing platform capable of mass-producing 6G antennas in a manner similar to semiconductor devices. Because the positions of the metal patterns and embedded switches can be determined together from the initial design stage, the geometry and array configuration can be tailored to the operating frequency and available installation space. The technology can also be extended to active antennas, transmit-array antennas, RIS, and other architectures. Further research increasing the number of switch paths could enable more precise, multilevel phase control.
With its minimal fabrication requirements, embedded architecture free of protruding components, low operating power, and ultrafast nanosecond-scale response, the technology is expected to reduce device size, weight, power consumption, and cost simultaneously. It could therefore serve as a foundational technology for the ultra-compact, ultra-low-cost mass production of beam-steering devices for low-latency 6G communications.
As the technology is further advanced, it could be used in base-station and satellite antennas for both terrestrial and non-terrestrial networks, as well as in thin beam-steering panels that supplement conventional repeaters or replace some of their functions. Potential applications extend further to Integrated Sensing and Communication (ISAC) systems that perform communications and sensing simultaneously on a single surface, communication control for fleets of robots and autonomous mobile systems in smart factories, low-Earth-orbit satellite connectivity, wireless power transfer, and high-speed wireless connections in AI data centers.
Professor Jungsuek Oh said, “This research moves beyond the conventional approach of fabricating an antenna first and then attaching tunable components. Instead, we designed and fabricated the semiconductor switches and antennas together within a single wafer from the outset. It is particularly significant that we were able to bridge the gap between semiconductor fabrication and radio-frequency systems by feeding the measured characteristics of the fabricated devices back into the antenna design.”
He added, “When large numbers of devices are formed simultaneously on a wafer, the assembly process and cost do not increase at the same rate as the array grows. Going forward, we plan to collaborate with foundries to scale the technology to large-area wafers and multilevel phase control, and to develop it into a versatile 6G radio platform applicable to terrestrial and satellite communications, ISAC, and industrial wireless systems.”
Jinhyun Kim, the first author of the paper, is currently a postdoctoral researcher at the University of California San Diego, where he conducts research on reconfigurable intelligent surfaces (RIS) and beam-steering system technologies. He plans to continue developing beam-steering systems through the integration of RFICs and RIS.
The research was supported by the Institute of Information & Communications Technology Planning & Evaluation (IITP), funded by the Korean Ministry of Science and ICT, under a program launched in 2021.
□ Introduction to the SNU College of Engineering
Seoul National University (SNU) founded in 1946 is the first national university in South Korea. The College of Engineering at SNU has worked tirelessly to achieve its goal of ‘fostering leaders for global industry and society.’ In 12 departments, 323 internationally recognized full-time professors lead the development of cutting-edge technology in South Korea and serving as a driving force for international development.
Nature Communications
Experimental study
Not applicable
The authors declare no competing interests.