Quantum key distribution (QKD) promises information theoretic secure key as long as the device performs as assumed in the theoretical model. One of the assumptions is an absence of information leakage about individual photon detection outcomes of the receiver unit. Here we investigate the information leakage from a QKD receiver due to photon emission caused by detection events in single-photon detectors (backflash). We test commercial silicon avalanche photodiodes and a photomultiplier tube, and find that the former emit backflashes. We study the spectral, timing and polarization characteristics of these backflash photons. We experimentally demonstrate on a free-space QKD receiver that an eavesdropper can distinguish which detector has clicked inside it, and thus acquire secret information. A set of countermeasures both in theory and on the physical devices are discussed.
Quantum key distribution (QKD) promises information theoretic secure key as long as the device performs as assumed in the theoretical model. One of the assumptions is an absence of information leakage about individual photon detection outcomes of the receiver unit. Here we investigate the information leakage from a QKD receiver due to photon emission caused by detection events in single-photon detectors (backflash). We test commercial silicon avalanche photodiodes and a photomultiplier tube, and find that the former emit backflashes. We study the spectral, timing and polarization characteristics of these backflash photons. We experimentally demonstrate on a free-space QKD receiver that an eavesdropper can distinguish which detector has clicked inside it, and thus acquire secret information. A set of countermeasures both in theory and on the physical devices are discussed.
Nowadays, the practical security of quantum key distribution (QKD) is the biggest challenge. In practical implementation, the security of a practical system strongly depends on its device implementation, and device defects will create security holes. The information leakage from a receiving unit due to secondary photon emission (backflash) is caused by a single-photon detector in the avalanche process. Now studies have shown that the backflash will leak the information about time and polarization and the eavesdropping behavior will not generate additional error rate in the communication process. An eavesdropping scheme obtaining time information by using backflash is proposed. Targeting this security hole for backflash leaking polarization information, an eavesdropping scheme for obtaining polarization information by using backflash is proposed in free-space QKD; however, it has not been reported in fiber QKD. In this study, the eavesdropping scheme and countermeasures for obtaining information by using backflash in fiber polarization-coded QKD is proposed. Since the polarization state of the fiber polarization-coded QKD system is easy to change, the scheme is proposed based on the time-division multiplexing polarization compensation fiber polarization-coded QKD system. In theory, the eavesdropper in this scheme obtaining the key information by using the backflash is theoretically deduced, and corrects the polarization change of the backflash by time-division multiplexing polarization compensation method, thus obtaining the accurate polarization information. The probability of backflash in the fiber polarization-coded QKD is measured to be 0.05, and the information leakage in the proposed eavesdropping scheme is quantified. The lower limit of the information obtained by the eavesdropper is 2.5 × 10–4. Due to the fact that the polarization compensation process increases invalid information in actual operation, the information obtained by the eavesdropper will be further reduced, thus obtaining the lower limit of information leakage. The results show that the backflash leaks a small amount of key information in a time-multiplexed polarization-compensated fiber polarization-coded QKD system. The wavelength characteristics of the backflash can be utilized to take corresponding defense methods. Backflash has a wide spectral range, and the count of backflash has a peak wavelength. So, tunable filters and isolators can be used to reduce backflash leakage, thereby reducing the information leakage.
Secondary photons released from photon detector when quantum cryptographic keys are shared between parties represent a significant eavesdropping threat. Attempts to listen in on quantum key distributions are, in theory, detected as automatic changes to quantum states. But Alice Meda from the INRIM in Italy and colleagues have now demonstrated that the detectors used to spot single photons in many commercial quantum key distribution boxes can leak information through ‘backflash’–bursts of light triggered when a photon hits an avalanche photodiode. The team constructed a sensitive new machine for detecting backflashes and determined data leakage rates nearing 10%. As well, each photodetector emitted specific signatures that might give attackers another entry point. Wide-bandwidth filters or special circulators incorporated into the quantum boxes can help secure this breach, the researchers propose.
InGaAs single-photon avalanche photodiodes (APDs) are key enablers for high-bit rate quantum key distribution. However, the deviation of such detectors from ideal models can open side-channels for an eavesdropper, Eve, to exploit. The phenomenon of backflashes, whereby APDs reemit photons after detecting a photon, gives Eve the opportunity to passively learn the information carried by the detected photon without the need to actively interact with the legitimate receiver, Bob. While this has been observed in slow-gated detectors, it has not been investigated in fast-gated APDs where it has been posited that this effect would be lessened. Here, we perform the first experiment to characterize the security threat that backflashes provide in a GHz-gated self-differencing APD using the metric of information leakage. We find that, indeed, the information leakage is lower than that reported for slower-gated detectors, and we show that its effect on the secure key rate is negligible. We also relate the rate of backflash events to the APD dark current, thereby suggesting that their origin is the InP multiplication region in the APD.
Quantum key distribution (QKD)1 is a quantum technology already present in the market. This technology will become an essential point to secure our communication systems and infrastructure when today’s public key cryptography will be broken by either a mathematical algorithm or by, eventually, the development of quantum computers. One of the main task of quantum metrology and standardization in the next future is ensuring that QKD apparatus works as expected, and appropriate countermeasures against quantum hacking are taken. In this paper, we discuss the security of one of the QKD most critical (and quantum-hackered) components, i.e., single photon detectors based on fiber-pigtailed InGaAs SPADs. We analyze their secondary photon emission (backflash light) that can be exploited by an eavesdropper (Eve) to gain information without introducing errors in the key. We observed a significant light leakage from the detection event of fiber-pigtailed InGaAs SPADs. This may represent a significant security threat in all QKD apparatus. We provide a method to quantify the amount of potential information leakage, and we propose a solution to fix this potential security bug in practical QKD apparatus.
Quantum key distribution (QKD) provides information-theoretic security grounded in the fundamental laws of physics. Nevertheless, practical imperfections can introduce side channels that expose QKD systems to quantum hacking, especially passive attacks that are inherently difficult to detect. In this study, we experimentally and theoretically investigate the upper limit of the backflash attack-a representative passive side-channel threat. Using a fully equipped fiber-based QKD receiver, we demonstrate the feasibility of the attack and reveal its limited capability in distinguishing quantum states. We further develop a theoretical framework to quantify the maximum distinguishability achievable by an eavesdropper, taking into account the broadband spectral nature of backflash photons. The analysis shows that Eve can extract effective key information from at most 95.7% of the backflash photons. Based on these findings, we evaluate the secure key rate of a decoy-state BB84 QKD system under backflash attack. Our results provide a quantitative assessment of the vulnerability of QKD systems to backflash emissions and offer a general methodology to evaluate the practical security of QKD systems.
In this article, we experimentally demonstrate an eavesdropper's (Eve's) information gain by exploiting the breakdown flash generated by the single photon avalanche detector (SPAD) used in coherent one-way quantum key distribution (COW-QKD) setup. Unlike prior studies focusing on the device-level characterization of backflash photons, this work quantifies Eve's learning with a QKD system that includes a key distillation engine (KDE). Eve's learning is quantified using the backflash photons emitted by SPAD and the information available on the classical channel. Experimentally observed data are in good agreement with theoretical simulations. Some mitigation strategies against the backflash attack are also discussed.