In our information society, the synthesis, distribution, and processing of radio and microwave signals are ubiquitous in wireless networks, telecommunications, and radars. The current tendency is to use carriers in higher frequency bands, especially with looming bandwidth bottlenecks due to demands for e.g. 5G and the “Internet of Things.” “Microwave photonics,” a combination of microwave engineering and optoelectronics, might offer a solution.
Electronic oscillators lie at the heart of virtually all microelectronic systems, generating the clock signals used in digital electronics and the precise frequencies that enable radio frequency (RF) sensors and communications. While an ideal oscillator provides a perfect signal at a single frequency, imperfections degrade the spectral purity of real-world components. Such impairments, broadly quantified as phase noise, ultimately limit the performance of many military radars and commercial 5G systems. The issue is becoming increasingly burdensome as the airways become more congested and defense needs evolve.
Today’s best microwave oscillators can achieve extraordinarily low phase noise, but the highest-performing technologies make large sacrifices in pursuit of performance. Trade-offs lead to oscillator modules with undesirable size, weight, power, and cost (SWaP-C), limited tunability, and high sensitivity to their surroundings, all of which limit their use in advanced defense systems.
The Generating RF with Photonic Oscillators for Low Noise (GRYPHON) program seeks to eliminate the shortcomings of today’s microwave oscillators by developing ultra-low-noise versions that are simultaneously compact, widely tunable, robust, and volume-manufacturable. To achieve its objectives, GRYPHON will employ emerging innovations in optical frequency division, integrated photonics, and non-linear optics.
Recent benchtop demonstrations using laser-based techniques have set world records in microwave phase noise. In parallel, ongoing innovation in the fields of integrated photonics and non-linear optics has enabled dramatic reductions in the size, weight, and power (SWaP) of key components needed to implement photonic oscillators. This includes chip-scale laser resonators with high quality factors and optical frequency combs.
A key building block of microwave photonics is optical frequency combs, which provide hundreds of equidistant and mutually coherent laser lines. They are ultrashort optical pulses emitted with a stable repetition rate that corresponds precisely to the frequency spacing of comb lines. The photodetection of the pulses produces a microwave carrier.
In recent years there has been significant progress on chip-scale frequency combs generated from nonlinear microresonators driven by continuous-wave lasers. These frequency combs rely on the formation of dissipative Kerr solitons, which are ultrashort coherent light pulses circulating inside optical microresonators. Because of this, these frequency combs are commonly called “soliton microcombs.”
Generating soliton microcombs needs nonlinear microresonators, and these can be directly built on-chip using CMOS nanofabrication technology. The co-integration with electronic circuitry and integrated lasers paves the path to comb miniaturization, allowing a host of applications in metrology, spectroscopy and communications.
Photonic microwave generation using on-chip optical frequency combs
Publishing in Nature Photonics, an EPFL research team led by Tobias J. Kippenberg has now demonstrated integrated soliton microcombs with repetition rates as low as 10 GHz. This was achieved by significantly lowering the optical losses of integrated photonic waveguides based on silicon nitride, a material already used in CMOS micro-electronic circuits, and which has also been used in the last decade to build photonic integrated circuits that guide laser light on-chip.
The scientists were able to manufacture silicon nitride waveguides with the lowest loss in any photonic integrated circuit. Using this technology, the generated coherent soliton pulses have repetition rates in both the microwave K- (~20 GHz, used in 5G) and X-band (~10 GHz, used in radars).
The resulting microwave signals feature phase noise properties on par with or even lower than commercial electronic microwave synthesizers. The demonstration of integrated soliton microcombs at microwave repetition rates bridges the fields of integrated photonics, nonlinear optics and microwave photonics.
The EPFL team achieved a level of optical losses low enough to allow light to propagate nearly 1 meter in a waveguide that is only 1 micrometer in diameter -100 times smaller than that a human hair. This loss level is still more than three orders of magnitude higher than the value in optical fibers, but represents the lowest loss in any tightly confining waveguide for integrated nonlinear photonics to date.
Such low loss is the result of a new manufacturing process developed by EPFL scientists — the “silicon nitride photonic Damascene process.” “This process, when carried out using deep-ultraviolet stepper lithography, gives truly spectacular performance in terms of low loss, which is not attainable using conventional nanofabrication techniques,” says Junqiu Liu, the paper’s first author who also lead the fabrication of silicon nitride nanophotonic chips at EPFL’s Center of MicroNanoTechnology (CMi). “These microcombs, and their microwave signals, could be critical elements for building fully integrated low-noise microwave oscillators for future architectures of radars and information networks.”
The EPFL team is already working with collaborators in US to develop hybrid-integrated soliton microcomb modules that combine chip-scale semiconductor lasers. These highly compact microcombs can impact many applications, e.g. transceivers in datacenters, LiDAR, compact optical atomic clocks, optical coherence tomography, microwave photonics, and spectroscopy.
DARPA GRYPHON program
“By implementing advances in photonic microwave generation with integrated photonics, we see a pathway to create a significant leap in microwave oscillator capability, while simultaneously realizing characteristics not found in today’s products: very low phase noise, compact form factor, ultra-wideband tuning, and environmental robustness,” says Gordon Keeler, DARPA MTO program manager. “Through GRYPHON, we hope to realize a major increase in capability for next-generation radar and communications systems.”
GRYPHON will explore innovative microwave sources using state-of-the-art microfabricated photonic components to achieve the target program metrics while creating a path to manufacturability. To accomplish the target objectives, the program will focus on two specific research areas. The first aims to develop a prototype that can be readily tested within an application and brought to maturation quickly. During the first phase of the program, research teams will prioritize achieving low phase noise and compact form factor, while tuning and robustness will be emphasized in later phases. The second research area will prioritize understanding the fundamental limits of photonic microwave generation. Research teams will be asked to offer at least an order of magnitude leap in one of three target metrics: size, phase noise, or frequency span.