Researchers at the National Institute of Standards and Technology (NIST) have devised and demonstrated the output components of a novel, quantum-based, self-calibrating standard for testing components and instruments in next-gen telecommunications networks. With further development, the system may eventually provide reference signals for networks running at, and soon far above, the present 5G range that can reach 24 to 39 billion cycles per second (gigahertz, GHz). That’s more than 10 times faster than 4G.
At this time, there is no quantum-based reference standard. As a result, it is already extremely difficult to measure, characterize, and calibrate signals accurately at 5G speeds, detecting problems such as waveform distortion and synchronization errors in components and systems. Eventually, “high-band” networks will operate at frequencies up to 100 GHz and possibly beyond, posing a formidable measurement challenge.
Moreover, at higher frequencies targeted for urban areas, waveforms lose their strength over shorter distances, so the fastest signals will have to be boosted more often by larger numbers of precisely synchronized cell sites and repeaters, all without altering the timing and shape of the waveforms. Monitoring and maintaining the integrity of such networks will demand measurement instruments and signal analyzers with dramatically increased capability, calibrated against an authoritative reference-signal standard.
“There are no primary reference-signal standards at these frequencies,” said NIST project scientist Pete Hopkins. “What is primarily used now are receivers or samplers that measure the power of an incoming signal. A reference source, or transmitter if you like, at 5G wireless frequencies, would be a game-changer.”
The new reference waveform source design, the first signal source with accuracy based upon quantum effects, is ultimately intended for deployment on a single chip. The work is described in the paper in IEEE Transactions on Superconductivity published on Feb. 3, 2020. It is based on previous groundbreaking work by NIST but exploits a different quantum technology.
The new design was inspired by NIST’s programmable Josephson Arbitrary Waveform Synthesizer (JAWS), the NIST Standard Reference Instrument for ac voltage. JAWS utilizes hundreds of thousands of synchronized microscopic superconducting devices called Josephson junctions (JJs) cooled to 4 kelvin (-269 ˚C), each of which generates a stream of accurate voltage pulses resulting from their intrinsic quantum behavior. Those pulses are combined to create a single larger voltage signal and generated in complex patterns to produce complex signals useful for communications measurements and calibrations.
A recent high-speed modification to the JAWS design, now in development, is able to reach frequencies of a few gigahertz – fast, but still much lower than necessary to serve as a calibration tool for the full range of 5G signal frequencies. That’s where the next steps come in.
In the new JAWS design described in the IEEE article, the researchers employ “single flux quantum” (SFQ) technology. Extremely tiny and brief electrical output pulses are emitted from Josephson junctions when they are excited by an electrical signal. Like many properties at extremely small scales, these tiny electrical pulses are quantized – that is, they can only take on exact, specific values, the smallest possible amount of which is a single flux quantum. And because they are quantized, their values are precisely known and controllable, the key to making a reference standard.
The pulse sequence patterns of the SFQs – in which the digital ones and zeros of the signal are represented by the presence or absence of an SFQ – are moved into a circuit loop that acts somewhat like a memory buffer in ordinary computers.
Once a series of SFQs is stored in the buffer – which is still in development – a high-speed clock shuttles the pulse stream out of the memory to the next stage. The speed and stability of this clock, which can operate at greater than 100 GHz – over 20 times faster than the clock in a typical PC – is built from SFQ technology and is the reason why this new reference waveform source can work at 5G frequencies.
Although this SFQ pulse stream is now fast, the individual pulses are too small to produce a useful signal. An amplifier is required to boost this signal. But this cannot be an ordinary amplifier. To serve as an accurate standard source, it must exactly multiply the signal while precisely combining and synchronizing the single flux quanta.
To do this, each SFQ pulse is routed to a “splitter” that produces multiple copies of the signal on branching channels. The amount of the multiplication depends on the layout of the branching channels. The present NIST design multiplies a single input into two, each of those into two more, and each of those into two more, resulting in an eightfold multiplication. Higher multiplications will be required before such a system could be deployed.
“The splitter works by duplicating the pulse multiple times,” said Manuel Castellanos-Beltran, first author of the IEEE report. “The extra energy required to do this is provided electrically to the components of the splitter tree.”
Finally, each of the multiplied pulses is routed to a series of superconducting quantum interference devices (SQUIDs). SQUIDs are exquisitely sensitive detectors of magnetic fields – such as the ones associated with the SFQ pulses – and are routinely used in medicine to detect and measure, for example, the exceedingly faint fields generated by nerve signals in the brain. An array of SQUIDs is used to combine the SFQ pulses because each SQUID can be inductively (magnetically) coupled to each of the splitter channels. The SQUID array adds all the duplicated pulses together to make one 8-fold larger pulse across the SQUID array. This signal is strong enough to be readily identified by electronic equipment.
“The availability of standard reference sources will enable signal quality improvements in all components in the 5G communications chain – transmitters, amplifiers, and receivers,” said Sam Benz, team leader for the project. “For example, problems with a receiver that add lots of noise and distortion can be easily identified if you know the source is providing the receiver with a clean and accurate signal.”
“We are excited to discover how far we can extend this technology,” Castellanos-Beltran said. “We need to demonstrate higher frequencies and higher output signal levels, both of which are challenging in their own right. That is what we are trying to figure out right now.”