CIR’s latest forecasts of the quantum repeater market are shown in the Exhibit below. The numbers come from “Quantum Networking Deployments, Components and Opportunities – 2017-2026,” a report that CIR published in October. Based on these forecasts, CIR has concluded that in about five years, revenues generated from quantum repeaters will represent a significant and attractive business opportunity.
Nonetheless, CIR believes that our forecast numbers understate just how important quantum repeaters are to the future of the Quantum Internet. Quantum repeaters should be seen above all as an enabling technology for the Quantum Internet. Without quantum repeaters revenues for quantum networking would be confined to the metro and local-area.
We have seen something like this before. Fiber optics found its way into the inter-switch segment of the telephone network in the very late 1970s but it was another five years before fiber optics made its way into the long-haul telephone network. And this couldn’t happen until the technology for erbium-doped fiber amplifiers (EDFAs) matured. As it went with fiber optic amplifiers in the 1980s, so it will go with quantum repeaters in the coming decade.
In looking at the R&D progress for quantum repeaters CIR sees the technology evolving through three main phases.
Packaging: First steps in quantum repeater commercialization: In China, extensive quantum networks have already been built use simple “trusted nodes” that measure and retransmit information about quantum states. But quantum repeaters at the present time are a long way from becoming standardized commercial products. What path will quantum repeater technology take to reach commercial products?
Now that the Quantum Internet is (1) receiving substantial government funding in China and Europe and (2) being driven by the arrival of commercialized quantum computers, CIR expects substantial progress towards repeater commercialization. The immediate prospects are for repeaters that facilitates plug-and-play placement in conventional long-haul optical networks.
There are already signs of this happening. In June 2017 SK Telecom said it had developed a quantum repeater that it will be installing in commercial networks. Sometime before this ZTE had already announced a QKD transport system for OTN networks, an extension of which would surely include quantum repeater technology. For the next two to five years, CIR expects to see a growing number of such announcements – Nokia is another likely player here. Such announcements position quantum repeaters more in the telecom mainstream, but don’t improve repeater functionality itself all that much.
More secure quantum repeaters: One significant limitation with current quantum repeater technology is that while it facilitates secure transactions through QKD, it may not itself be secure. A recent survey of quantum repeater vulnerabilities by Japanese scientists (http://wp.internetsociety.org/ndss/wp-content/uploads/sites/25/2017/09/01_2_3.pdf ) is quite devastating in this regard.
“Quantum” repeaters today are actually hybrid systems and include classical computing devices that – as the Japanese scientists point out – are just as vulnerable as other classical systems to security violations. It is true that a quantum repeater can be physically secured, but quantum encryption based on repeaters that are not themselves secure, detracts from the business case for QKD and could also be a problem for those sharing of quantum computer resources over a cloud and/or a network.
There is a growing research literature on what to do about all this, but very little likelihood that bit/qubit conversions inside the repeater will ever be wholly eliminated. Rather, what CIR anticipates is that enough will be done over the next five years to produce repeaters that end user concerns will be addressed. Repeater vendors will then declare victory and introduce a new wave of next-generation secure repeaters. Based on the current state of the R&D and the speed at which we think quantum networks will develop, we think this new breed of secure quantum repeaters will emerge in about three to five years.
The next best thing to a quantum amplifier: Unless someone finds a way round the non-cloning theorem in quantum theory, there will never be a quantum analog of the optical amplifier. However, what may be possible – and surely this will be more than five years from now – is a highly slimmed down quantum repeater that will be low-cost and easy to install in much the same way that EDFAs are.
If this stage in the evolution of the quantum repeater actually occurs, it will probably be a device that does not have extensive integrated memory requirements. At the present time, we are at a stage in quantum repeater development where storage is largely classical, raising the vulnerability issues mentioned above. The next stage will be quantum repeaters utilizing quantum memories. But NTT for one is researching quantum repeaters without quantum memories. Its optical quantum repeater, designed with some Canadian researchers, appears to eliminate the quantum memories in repeaters.
It is still too early to say where this work is headed from a commercial perspective. However, if this kind of quantum repeater ever becomes the commercial product that enables a great leap forward in the Quantum Internet, history of the EDFA and long-haul optical networks will have repeated itself. As in the Optical Internet, so in the Quantum Internet.