The squeeze is here, and the crunch is coming. Soon.
Explosive demand for high-speed wireless communication is placing growing pressure on the limited frequency spectrum bands allocated for mobile phones, Wi-Fi electronics, and related uses. Today’s typical cell phone operates from around 800 megahertz (MHz, millions of cycles per second) to 2 gigahertz (GHz, billions of cycles). That is in the same UHF range already occupied by broadcast TV, microwave ovens, GPS, Bluetooth, and two-way radios, among many other applications. There is no room left to accommodate expansion.
The most promising solution to this “spectrum crunch” is to drastically increase the frequencies at which future wireless devices, such as so-called “5G” phones, can operate -- by an unprecedented factor of ten in the near term, and ultimately by more than a hundred. But at present there are no authoritative, quantum-accurate* standards that can enable manufacturers of next-gen wireless devices and operators of high-speed networks to verify that their equipment will actually perform correctly at such frequencies.
Researchers at the National Institute of Standards and Technology (NIST) have recently taken two major steps to rectify that situation. Earlier this year, they filed a preliminary patent application for an advanced device to provide quantum-accurate radio frequency (RF)** sources up to 1 GHz and beyond. And they have now reached a milestone in development of a novel alternate technology – pioneered at NIST – with the potential to reach 300 GHz.
“Future communications will need a standard that is not only linked to the International System of Units, but that is repeatable, programmable, and accurate all the time. That just doesn’t exist now,” said Samuel Benz who heads NIST’s Superconductive Electronics Group. “There are no components, no electronics, and no metrology yet to address the very stringent standards that will be needed to cover this wide frequency range.”
Both of NIST’s latest efforts exploit the quantum behavior of Josephson junctions (JJs) ***, which consist of two tiny superconducting electrodes separated by a very thin non-superconducting barrier. JJ devices, in somewhat different configurations, serve as the world’s de facto standard for direct-current (dc) and alternating-current (ac) voltage.
To produce exactly quantized dc voltages, an RF signal (typically in the tens of GHz range) is applied to the junction. The output is a series of quantum-exact dc voltage steps that depend only on the applied frequency and fundamental constants of nature. NIST researchers invented and have continuously improved a design with over 260,000 JJs that allows users to select specific accurate voltages. That technology is available as the Programmable Josephson Voltage Standard in NIST’s inventory of Standard Reference Instruments (SRIs).
The same general JJ design can also be used to generate alternating-current (ac) voltages at a range of frequencies. In a scheme originated at NIST, a programmed stream of electrical pulses is applied to the junction, which “locks” onto the applied pulses to produce ac waveforms – not only sine waves but other forms as well. The result, after many iterations and improvements, is NIST’s Josephson Arbitrary Waveform Synthesizer (JAWS), also an SRI.
Previously, circuit-based limitations compromised the quantum accuracy of the signals produced by JAWS for frequencies beyond about 20 kilohertz (kHz, thousands of cycles per second) – about the highest frequency that humans can hear, but far too low for telecommunications. NIST researchers have recently developed new circuits and bias techniques that reduce these errors and have the potential to generate quantum accurate signals up to a frequency of a few gigahertz.
However, there are several other technical limitations, including the lack of an adequately fast pulse generator. In addition, the system needs to output controllable and uniform waves dependably. The existing JAWS devices can’t do that optimally at high frequencies because the process of creating the crest and trough of each wave causes a timing shift that causes signal distortion and loss in accuracy of the voltage output.
The new RF JAWS design in the provisional patent application solves this problem by separating the JJs into two separate groups: One produces crests, and the other troughs. The parts are combined and synchronized into a train of complete, distortion-free waveforms.
“We expect that this design will get us to 1 GHz,” Benz said, “and we hope to deliver a standard within two years.”
Meanwhile, to generate quantum-accurate signals at 5G frequencies, NIST is pursuing an alternative solution by building faster superconducting digital circuits that use, coincidentally, JJs as the switching devices instead of semiconductor transistors. These superconducting circuits work by storing and manipulating tiny, very fast electrical pulses, called single flux quantum (SFQ) pulses.***
A flux quantum is essentially a stored persistent current in the superconducting loop. An incoming current pulse, together with the current already passing through the JJ, can cause the JJ to switch and the SFQ to be transferred to the next junction in the row. The final output is a voltage pulse. Its time-integrated voltage is exactly h/2e, where the Planck constant h**** and the electric charge of a single electron (or proton), e, are fundamental constants of nature.
Because the JJ switches are fast, and their output is exactly quantized, SFQ technology has the potential to extend the range of synthesized waveform frequencies with quantum-exact amplitudes into the tens or hundreds of gigahertz. Since the SFQ pulses are so small, the output voltage of each circuit is tiny, and thousands must be integrated to reach desired, much larger voltages.
“Recently, we’ve demonstrated 5 GHz at less than a millivolt, and we want to demonstrate 30 GHz at 10 millivolts,” said Peter Hopkins, who leads the Flux Quantum Electronics Project. “Industry would like to have all the way up to 100 millivolts or even 1 volt if possible.”
There will be numerous challenges. One involves fabrication of circuit components at vanishingly small dimensions. The JJs on the present voltage standards measure about 7 micrometers (µm, millionths of a meter); for the SFQ circuits, the JJ diameter is about 1 µm, and the researchers are working on 0.3 µm. The barriers between the superconducting electrodes of the Josephson junction are about 5 nanometers (nm, billionths of a meter) in thickness.
“What’s more, this program is pushing the forefront of the technology to the point where there isn’t any commercially available equipment,” Benz said. “We have a lot of trouble just measuring our circuits at hundreds of gigahertz. But the stakes are high. Precision timing and control of single flux quanta at hundreds of gigahertz is essential for the next generation of communication devices and systems. In a few years, we hope to have calibration sources and a standard reference instrument for use by wireless chip and instrument manufacturers.”
* “Quantum-accurate” means that a measurement value is based on a fundamental effect of quantum mechanics that is always the same and therefore is not subject to error and is robust to changes in the environment.
** Radio frequency extends from about 20 kHz to 300 GHz. This includes, at the upper end, “millimeter waves” from 30 GHz to 300 GHz.
*** Magnetic flux is basically the amount of magnetic field that penetrates a surface. It can be graphically demonstrated by placing a sheet of paper over a magnet and sprinkling iron filings onto the paper. The resulting pattern shows the flux lines. At extremely small dimensions, flux is quantized – that is, it can take on only specific discrete values.
**** This constant is fundamental to modern science. E = hν is the first "quantum" expression in history, formulated by physicist Max Planck in 1900. Here, E is energy of a photon, ν is frequency, and h is what is now known as the Planck constant.