Cyber security is an increasingly complex issue and a growing concern worldwide. As our utilization of communication technology increases, so to do the means used by cyber intruders to infiltrate networks to access sensitive information or damage infrastructure. The cornerstone of secure communications is the cryptographic keys. As the varying heights of a door key’s cuts can keep a door locked, a varying numerical sequence – or random number sequence – can be used to keep a data message locked. Everything from secure email and online banking to keeping our modern power-grids and satellites safe relies on these data keys.
Traditionally, random numbers are computer-generated using mathematical algorithms. This generation technique can be problematic because computer calculation methods are inherently deterministic, meaning that the results produced are never truly random but rather pseudo-random. If an adversary determines the algorithm or even any biases in the algorithm, security can be compromised. A significant number of high-profile cyber compromises targeting random number generation have driven the search for more reliable sources of random numbers.
Perhaps what some may consider an unlikely source for achieving this, is light, specifically random number generation (RNG) using lasers to amplify quantum mechanical noise.
Along with a number of organizations around the world, the National Research Council of Canada (NRC) is working to understand how quantum mechanics can be used to build novel devices and processes not possible with conventional physics. These quantum technologies will bring a host of new opportunities in understanding fundamental physics and will drive new high-technology applications in sensing, communications, computing and beyond in the coming decades.
Over the years sources of randomness have been investigated, including radioactive decay, atmospheric fluctuations, and even the mysterious motions of lava lamps! Quantum mechanical processes are now generally established as a preferred source of randomness. There is no algorithm that can predict the outcome of a probabilistic quantum mechanical process, unlike with the previously discussed computer-based techniques. With quantum mechanical experiments it is not just that we don’t know what the outcome will be, but even more so it’s that the outcome is unknowable in advance. This unpredictability is what makes quantum mechanics so unique and powerful for the RNG application.
Using intense lasers we can convert vacuum fluctuations, or quantum mechanical wiggles that occur as a result of Heisenberg’s uncertainty principle, into light and then amplify them so that they can be easily measured. We can compare this method of RNG to throwing quantum dice. The outcome of a quantum measurement can be used to generate a number, only the quantum uncertainty of the roll renders the outcome unpredictable. In contrast, a careful study of the classical motion of dice spinning through air – perhaps by video capture and classical analysis – would allow the prediction of which side will land up when it hits the ground.
Photonic methods are particularly effective for rapid generation because of both practical and fundamental reasons. Practically, the sophistication and accessibility of photonic technology, a result of a confluence of an explosion in optical data network hardware and advances in research, technology, and manufacturing, allow high-speed systems to be developed at increasingly lower costs. Fundamentally, photons and light-matter interactions often lie at the heart of quantum mechanical behaviours so it’s perhaps not surprising that the manifestation of uncertainty is apparent in many optical systems.
Our increasing use of secure information networks, ranging from online storage to high-bandwidth communications, requires a need to constantly push the limits of what is technologically available. As network speeds increase, speed of RNG can become an increasingly critical parameter. With this in mind, since NRC’s original implementation of a quantum key generator, we have worked to reduce the power requirements of laser-induced quantum RNGs to ensure that we can repeat the sequences at high-speeds, something we were not able to do in our earliest implementation. Other improvements have come in the form of the type of vacuum fluctuations that we can measure. For example, the first devices we built looked at randomness in the phase of the amplified light but we have recently developed an alternative approach where we measure fluctuations in the brightness, or pulse energy, of the light. We can now generate high-performance random number sequences with low power requirements. Sometimes speed isn’t the only important feature for key generation and we’re looking at other practical requirements too, including robustness and fail-safe operation. The features in these prototypes strengthen existing secure communications systems, making quantum augmented networks, and ultimately they will be used in completely secure quantum communications systems of the future.
Ben Sussman leads the Quantum Photonic Sensing and Security Program at the National Research Council of Canada, a 40 person effort focused on developing next-generation photonic technologies. He is also adjunct professor of Physics at the University of Ottawa. His research investigates building quantum technologies at the intersection of ultrafast quantum control and quantum optics.