Airborne demonstration of a quantum key distribution receiver payload

1. Introduction

• Terrestrial fiber and free-space QKD are limited to a few hundred kilometers by losses and, for free space, atmospheric conditions and line-of-sight requirements.Fiber absorption losses scale exponentially with distance; free-space geometric losses scale quadratically.Source: §1 · p. 2
• Satellite links could support global QKD networks, and the Canadian QEYSSat concept proposes an uplink with the receiver on the satellite and source on the ground.Other satellite mission analyses primarily propose downlinks, with photons generated in orbit and sent to ground receivers.Source: §1 · p. 2
• Uplink QKD offers simpler satellite design, relaxed processing and storage requirements, and source flexibility, although comparative analysis found fewer key bits than downlink by less than one order of magnitude.Under similar conditions, atmospheric turbulence affects the uplink beam path earlier in propagation.Source: §1 · p. 2
• The paper demonstrates QKD from a ground transmitter to an aircraft receiver, addressing the absence of reported aircraft uplink demonstrations and operating at angular rates exhibited by LEO satellites.The receiver uses components designed for satellite mass, volume, power, thermal, and vacuum constraints; many are already space-suitable or have a clear path to flight.Source: §1 · p. 3

2.1. Concept

• Two night flights comprised 14 passes at nominal distances of 3, 5, 7, and 10 km, using arc and straight-line paths; tracking feedback and beacon lasers guided the two-axis motors.The links were attempted at night to limit optical noise, and the aircraft flew at about 1.6 km altitude.Source: §2.1 · pp. 3–4, Figure 1 · p. 4

2.2. Source and Transmitter

• The ground source implements 400 MHz BB84 with decoy states, selecting signal, decoy, or vacuum pulses and preparing the four BB84 polarizations.The state sequence repeats every 1,000 pulses; the authors note this is insecure but sufficient for the demonstration and that upgrading to a fully random sequence is straightforward.Source: §2.2 · pp. 4–5, Figure 2 · p. 5

2.3. Receiver

• The aircraft receiver collects the signal with a 10 cm-aperture telescope, then filters and analyzes polarization into four BB84 measurement states for detection and time-tagging.Its custom receiver includes a fine-pointing unit and integrated optical assembly, detector module, and control and data processing unit; the apparatus was mounted facing out the aircraft cabin door.Source: §2.3 · pp. 6–7, Figure 3 · p. 7

2.4. Acquisition and Calibration

• Beacon cameras and two-axis motors provide coarse pointing, while inertial navigation and a wide-divergence infrared light support initial acquisition before two-way beacon tracking.Independent beacon-to-signal alignment and camera calibration were performed before each flight, allowing link acquisition to begin when the aircraft reached the ground station vicinity.Source: §2.4 · pp. 8–9

• 6 of 7 successful-link passes yielded secret key, out of seven successful quantum links among 14 aircraft passes.Source: §3 · p. 9
• At 3 km on an arc pass, maximum measured angular rate was 1.28°/s, consistent with the cited LEO rates of 0.72°/s and 1.2°/s.Source: §3 · p. 9
• Arc passes supported longer key exchange, while straight-line passes more closely represented a satellite pass’s angular-speed change.Source: §3 · p. 9
• Across successful passes, mean typical pointing error ranged from 0.00133° to 0.0220° at the transmitter and 0.0630° to 0.126° at the receiver.Source: §3 · p. 9
• 34.4–51.1 dB was the mean measured quantum-link loss; modeled estimates were lower bounds under assumed diffraction-limited divergence, and intentional defocus made observed losses generally higher.Source: §3 · p. 11
• After the SNR filter, receiver QBER for secure-key passes ranged from 2.96% to 5.24%.Source: §3 · p. 12
• Five of the six passes eligible for key extraction yielded secure keys after finite-size effects; the remaining pass did so only when finite-size effects were ignored.Source: §3 · p. 12

4. Discussion and Conclusion

• With source-intrinsic QBER typically 2%–4%, finite-size secure keys were extracted on many passes while the aircraft moved at up to 259 km/h.The optical links spanned 3–10 km, and BB84 decoy-state signals were sent at angular speeds similar to those of a low-Earth-orbit satellite.Source: §4 · p. 12
• Space use still requires component-specific modifications, including radiation-hard CDPU parts and low-outgassing IOA glues.Proton radiation can increase Si-APD dark counts; cooling and thermal or laser annealing can mitigate these effects, and a space-suitable prototype detector module is being developed.Source: §4 · pp. 12–13
• The receiver components were shown in a concept placing them on the flight-proven NEMO-150 micro-satellite bus, supporting the feasibility of an uplink QKD satellite mission.The figure shows the bus externally with its extended telescope and baffle, and internally with possible placement of the demonstrated receiver components.Source: Figure 5 · p. 13, §4 · p. 13
• Ground-to-satellite operation may require predicted-position pointing, point-ahead, and fine pointing over longer distances; these were unnecessary for the aircraft.Point-ahead depends on the transmitter’s divergence and compensates for satellite motion during optical-signal flight time.Source: §4 · p. 13
• The receiver is expected to support entangled-photon-pair QKD using a BBM92-style protocol without receiver changes.The proposed arrangement measures one photon of each pair on the ground.Source: §4 · p. 13