In recent years, quantum computers have made headlines, while quantum communication solutions have made it to markets. Simple solutions, such as quantum random number generators, have been transferred into smartphones. Similar quantum encryption modules for terrestrial telecom router systems are available off-the-shelf. Courtesy of Fraunhofer IOF. But this is just the beginning, and more sophisticated solutions are in the development pipeline. “[Quantum communication] adoption is expected to accelerate over the coming decade as quantum computing cybersecurity risks increase,” according to McKinsey & Company’s third annual Quantum Technology Monitor, in April 2024. Quantum cryptography basics Quantum communication essentially refers to quantum encryption. And essentially, for quantum encryption, two quantum effects are exploited. The simpler of the two is based on the statistical nature of quantum effects. If, for example, the radiation of a laser is attenuated until there are one or zero photons per time unit left remaining, then this light signal can be exploited as a random generator. This can be assumed to be physically safe, since such light emission is unpredictable by its quantum nature. A sequence of photons (or signals with several photons) and voids can easily be converted into random numbers. Other quantum effects can also complete this function. Entanglement is a more sophisticated effect. To understand entanglement, it is critical to know that some quantum effects produce two photons that have a defined relation, such as perpendicular polarization states. The peculiar aspect of this quantum effect is that the exact states of both photons remain undefined until the state of one of the two photons is measured. If one state is measured, the other gets fixed instantaneously. The 2022 Nobel Prize in physics was awarded to three scientists who proved this behavior experimentally. As for cryptography, the benefit appears simple: An entangled photon can be measured only once. Effectively, eavesdropping by a third, malicious party to observe this effect becomes impossible, because it destroys the entanglement. It is also important to note that other quantum systems, with different physics, and sometimes more states, show a similar behavior. The problem with entangled photons (or any other entangled quantum entity) is its elusive nature. Measuring single photons has been performed in microscopy long ago, for example, but noise remains a challenge in this case. The generation of single photons is even more difficult, particularly as it relates to heralding: As a pair of entangled photons is destroyed by any measurement, a separate heralding signal is helpful for use as a trigger of subsequent processes. Often, a unification of the two quantum communication ideas can solve this heralding problem. When entangled photons are generated, one may serve as a herald, or trigger signal, while the other serves as part of a random sequence that becomes the actual quantum key. It is also important that single photons can neither be amplified nor replicated; one remains one. Replicating a single photon into a similar “fresh” photon does not change anything. Fractional attenuation is impossible and thus, the photon is either there or it is lost. Therefore, the distance over which a quantum signal can be transmitted (i.e., single photons) depends on the attenuation of that line. In conventional fibers, this value is up to ~80 km, and in free space, it might extend to thousands of kilometers if there is no air in between. Repeaters that reproduce entanglement are a subject of considerable research efforts, as are entanglement and/or quantum storage devices. Quantum key generation is vital in quantum communication, since, in most applications, it is seen as sufficiently secure if a quantum key is exchanged and if larger data volumes can be exchanged afterward, using the key in conventional encryption algorithms. In other words, the transmission of secret data could be secure if entangled photons were used. However, these entangled photons are difficult to generate, suffer from high transmission losses, and are subject to detection noise. As a result, the rate of successful transmission of entangled signals is low, often on the order of a few bits per second. No matter if the signal is an entangled photon or a quantum random signal, the secure quantum signal is usually used only for the transmission of the quantum key. After a successful quantum key transmission, the real data is encrypted and transmitted on a conventional connection. Quantum key distribution The method of preparing random quantum states and sending them — for example, from Alice to Bob — is referred to as prepare and measure. Sophisticated protocols have been developed to ensure that the quantum key sequence is transmitted and processed safely and efficiently. The most common protocol is known as BB84, as it was proposed by Charles Bennett and Gilles Brassard in 1984. This protocol uses the polarization of photons as the quantum effect for a safe quantum key transmission (or transmissions). Alice sends a sequence of polarized photons to Bob, where the base can switch between horizontal/vertical (H/V) and diagonal/antidiagonal (D/A). To exclude eavesdroppers, the actual base is fixed after the transmission and half of the transmitted data is discarded. Part of the data is compared between Alice and Bob and the rest remains a safely transmitted quantum key. Rather than relying on entangled photons, this protocol relies only on a sequence of randomly generated quantum states, from a trusted source, which serve as a cryptographic key. These states may consist of several photons. While this makes the generation and detection of polarization states easier, it also allows an eavesdropper to strip off single photons from the signal. These are called photon-number-splitting attacks. The practice of using decoy states can widely suppress this class of attacks. Decoy states are additional photon pulses with varying intensities inserted alongside the actual signal pulses used to encode key information. Decoy states are used to capture photon statistics. Since eavesdroppers cannot distinguish between signal and decoy pulses, any attempt to interfere with the communication will be detectable through the discrepancies introduced in the decoy statistics. Alice and Bob can compare these statistics to determine whether photons were stripped off. In such a case, they would repeat the measurement until the statistics are equal. A photon source for quantum states When Bennett and Brassard presented their protocol for safe quantum communication in 1984, they assumed a weak, or faint, photon source1,2. For commercial purposes, this is still the most popular way to generate signals for quantum communication. The primary challenges are in creating a photon source that can produce indistinguishable and randomly polarized photons at a high rate while maintaining compactness and energy efficiency. In 2014, Gwenaelle Vest and Harald Weinfurter proposed the use of VCSELs for this purpose3. This solution enabled high pulse rates in the gigahertz range, and it made the technology both compact and field-deployable. Now, a team at Fraunhofer Institute for Applied Optics and Precision Engineering IOF (Fraunhofer IOF) in Jena, Germany, has developed a new source based on the VCSEL concept. The Fraunhofer IOF researchers’ hybrid faint pulse source uses a linear array of eight VCSELs, promising higher spectral and temporal indistinguishability, enhanced polarization quality, and robust scalability for real-world applications (Figure 1). The system was in fact developed for a potential communication link from a low-Earth-orbit satellite to an optical ground station. Figure 1. A schematic of the eight-channel VCSEL source for polarization encrypted photons. See Reference 4. Courtesy of Fraunhofer IOF. Backbone of the photon source The developed system uses a common gallium arsenide (GaAs) substrate for eight VCSELs with lithographically structured polarizers developed at the University of Stuttgart to enable the precise generation of polarization-encoded photons. This allows the source to efficiently support the four polarization states required by the BB84 protocol (H/V/D/A), while maintaining its ultracompact form factor. One of the main features of the developed system is its spectral indistinguishability, which it achieves through temperature leveling of the substrate to maintain a variation of significantly less than 0.5 K across the array. Centered at 850 nm, the wavelength differences are therefore approximately 90% overlap at full width at half maximum. Further, the integrated polarization gratings are electron beam written and can be aligned very precisely to the required direction. Preliminary data showed that on-chip polarizers can achieve at least 12-dB extinction ratio in diagonal, and at least 20 dB in a horizontal or vertical direction4. System architecture and optical design Four of the eight VCSEL channels are designated to supply decoy states by using an attenuator that reduces the amplitude of these pulses by ~4 dB. This approach enhances the overall security of the quantum communication link, as it allows the system to produce both signal and decoy pulses at the same spectral and temporal conditions. An integrated digital-to-analog conversion further enables the lasers to be driven with up to 5 GHz (Figure 2). It is expected that this signal is to come from an additional quantum random number generator. Figure 2. A ceramic printed circuit board (PCB) is bonded to a molybdenum heatsink of a similar size. The PCB carries the VCSEL and driver chip (center). The wings are plugs for thermal management components. The plug in the back will connect to the gigahertz driver signal. See Reference 4. Courtesy of TU Ilmenau. The optical system of the source is housed in a KOVAR box with a low expansion coefficient (Figure 3). It houses the fixtures for a first collimation microlens array, the attenuator for the decoy states, and another microlens array for refocusing the eight beams onto the waveguide where they are merged into one output channel (Figure 3). The overall coupling efficiency is 99% without Fresnel losses. The 1% loss is due to the residual elliptical mode mismatch when the VCSEL modes are coupled into the waveguide, which is surrounded by a heatsink, which itself warrants a closed-loop temperature control to avoid any additional birefringence. The polarization-independent waveguide combiner is provided by the Institute of Photonics and Nanotechnology at Politecnico di Milano in Italy. Figure 3. The printed circuit board (PCB) in Figure 2 is shown with the KOVAR frame housing the optics that are fixed on top of the VCSEL chip. The tiny glass tip on top of the housing is the waveguide combiner, where the polarization signal comes to exit. Courtesy of Fraunhofer IOF. The eight-channel VCSEL source for BB84-based quantum key distribution with decoy states is miniaturized with integrated components using a space These specifications make the solution a strong candidate for a cubesat-size space mission, and all technologies have been selected to be ready for space qualification in the future. The prototype will be presented at Photonics West 2025. Meet the authors Erik Beckert earned his diploma in precision engineering and a doctorate in optoelectronics system integration from Technical University of Ilmenau, Germany. He heads the Opto-Mechatronical Components and Systems department at Fraunhofer IOF in Jena, Germany; email: erik.beckert@iof.fraunhofer.de. Christopher Spiess is a research associate and photonics innovator at Fraunhofer IOF. As a senior system engineer and project team leader, he specializes in advanced quantum communication systems and their technological implementations. Spiess earned a Ph.D. from Friedrich-Schiller University Jena; email: christopher.spiess@iof.fraunhofer.de. Andreas Thoss, Ph.D., is a laser physicist, founder of THOSS Media, and a contributing editor to Photonics Spectra. He has been writing and editing technical texts, with a focus on the field of photonics, for two decades. References 1. C. Bennett and G. Brassard (December 1984). Quantum cryptography: Public key distribution and coin tossing. Proc IEEE Int Conf Comput Syst Signal Process, pp. 175-179, Bangalore, India. 2. T. Mor and R. Renner. (December 2014). Theoretical Aspects of Quantum Cryptography — celebrating 30 years of BB84. Theor Comput Sci, Vol. 560, No. 1. 3. G. Vest et al. (2014). Design and evaluation of a handheld quantum key distribution sender module. IEEE J Sel Top in Quantum Electron, Vol. 21, No. 3, pp. 131-137. 4. E. Beckert et al. (March 2024). Ultra-bright polarization entangled photon sources, e.g. based on time-reversed Hong-Ou-Mandel interference. Proc SPIE 12911, Quantum Computing, Communication, and Simulation IV 129110U, San Francisco, California.