Realistic Threat Models for Fiber and Free-Space Continuous-Variable Quantum Key Distribution

TL;DR

The paper develops a realistic threat model for Gaussian-modulated coherent-state CV-QKD with an LLO, combining composable finite-size security with a multi-level trust framework for loss and noise in both fiber and free-space channels, including satellite links. By deriving asymptotic and finite-size key-rate expressions and mapping how Eve’s capabilities reshape the covariance structure, it shows that trusted transmitter and receiver noises can substantially raise achievable rates, while untrusted transmitter loss remains a critical bottleneck. The analysis demonstrates that satellite-to-ground CV-QKD, under LoS constraints, can outperform ground-based repeaterless bounds and rival repeater-assisted links, illustrating the potential for a global high-rate quantum network using current technology. Overall, the results emphasize mitigating source leakage and phase noise and leveraging LoS-enabled satellites to realize robust quantum-safe communications at scale.

Abstract

Future global quantum communication networks, or quantum Internet, will realize high-rate secure communication and entanglement distribution for large-scale users over long distances. Continuous variable (CV) quantum key distribution (QKD) provides a powerful setting for secure quantum communications, thanks to the use of room-temperature off-the-shelf optical devices and the potential to reach high rates. However, the achievable performance of CV-QKD protocols is fundamentally limited by the fact that they appear to be fragile to both loss and noise. In this study, we provide a general framework for analyzing the composable finite-size security of CV-QKD with Gaussian-modulated coherent-state protocol (GMCS) under various levels of trust for the loss and noise experienced by the users of the protocol. Our work is comprehensive of several practical scenarios, encompassing both active and passive eavesdropping configurations, with both wired (i.e., fiber-based) and wireless (i.e., free-space and satellite-based) quantum communication channels. Our numerical results evaluate the robustness of the GMCS protocol under varying levels of trust and demonstrate that it is difficult for a practical protocol to remain robust against untrusted loss at the transmitter. In the wireless case, we analyze a scenario with a sun-synchronous satellite, showing that its key distribution rate, even with the worst level of trust, can outperform a ground chain of ideal quantum repeaters. Our results indicate that, when it comes to engineering and optimizing quantum-safe networks, it is essential to mitigate the shortcomings caused by critical trade-offs between rate performance, trust level, system noise, and communication distance.

Figures (13)

• Figure 1: Quantum communication scenario between transmitter (Alice) and receiver (Bob) separated by a quantum channel with transmittance $\eta_{\rm ch}$ and mean number of input thermal photons ${\bar{n}_e}$. Alice’s setup has extra thermal photons ${\bar{n}}_{\rm Tx}$. Bob’s setup has detection efficiency $\eta_{\rm eff}$ and extra thermal photons ${\bar{n}}_{\rm Rx}$. Sig: signal laser; Mod: random Gaussian displacement; QRNG: quantum random-number generator. We also describe the various trust levels. Eve (1): Eve only attacks the quantum channel. Eve (2): Eve collects leakage from both the quantum channel and the receiver’s setup. Eve (3): besides the two leakages above, Eve also performs an active side-channel attack on Bob's setup, so that the noise internal to the setup is untrusted. Eve (4) and Eve (5) are enhanced versions of Eve (1) and Eve (3), respectively. In these cases, Eve can collect leakage from Alice's setup and control the Alice's setup noise.
• Figure 2: The collective Gaussian attack under the assumption of untrusted loss and noise in both Alice and Bob side, i.e., Eve (5) in Fig. \ref{['fig:theory']}. QM: quantum memory; ${\varphi _{{\rm{TMSV}}}}$: TMSV state.
• Figure 3: (a) The optical layout of the LLO scheme. AM: amplitude modulator; MBC: modulator bias controller; DAC: digital-to-analog converter; IQmod: in-phase and quadrature electro-optic modulator; PD: photo diode; BS: beam splitter; VOA: variable optical attenuator; PC: polarization controller; LO: local oscillator; ADC: analog-to-digital converter. (b) Structure of an optical IQmod. $u_I$, $u_Q$ denote the amplified voltages carrying the information. ${u_{0,I}}$ and ${u_{0,Q}}$ are the bias voltage of MZM1 and MZM2, respectively. MZM: Mach-Zehnder modulator. PM: phase modulator. (c) Schematic of a balanced homodyne detector. $+U_0$ and $-U_0$ denote the external linear power supply. ${g_e}$ is the electric amplification, and $U$ is the measurable voltage. TIA: transimpedance amplifier.
• Figure 4: Fiber-based CV-QKD system multiplexed with classical channels. OTU: optical transform unit; OMU: optical multiplexer unit; OA: optical amplifier; WBC: wide band coupler; SMF: Single-mode fiber.
• Figure 5: Intensity and wavelength distributions of Raman scattering and Brillouin scattering. Here, the pump wavelength is 1550 nm, which is located at C band.