TL;DR
This paper presents a novel, resource-efficient method for increasing the probability of delivering single photons on demand. By using a simple fiber-optic delay loop and a single switch, they effectively "stack" photons generated from four consecutive laser pulses into a single time bin, boosting the delivery probability without needing multiple complex sources. This is a crucial step towards scaling up quantum technologies that rely on synchronized single photons.
What Problem Does This Paper Address?
Quantum information processing and enhanced quantum technologies (like quantum computing and secure communication) critically depend on the ability to generate and deliver single photons reliably and on demand. A common method for generating single photons is using "heralded" sources, where the detection of one photon (the herald) signals the presence of its entangled twin. However, these sources are inherently probabilistic – they don't always produce a photon pair with each attempt.
To get more photons, researchers have explored multiplexing, which combines the outputs of multiple sources. While effective, traditional multiplexing schemes (like spatial or spectral) often require a large number of sources and complex setups, leading to significant resource overhead. This paper tackles the challenge of increasing single-photon delivery probability in a more resource-efficient way, specifically by temporal multiplexing of photons from a single source. The goal is to achieve higher photon delivery rates without a proportional increase in the number of physical sources.
Methods
The core of this work is a "temporal loop multiplexing" scheme. Here's how it works:
- Heralded Single Photon Source: They start with a well-established fiber-based source that generates heralded single photons via four-wave mixing in a photonic crystal fiber (PCF). This source is designed to produce high-purity photons.
- Fiber Delay Loop: A key innovation is the use of a simple, fully fiber-integrated delay loop. This loop is designed to introduce a delay matching the repetition rate of the laser pulses (200 ns in this case).
- Single Optical Switch: A single, switchable 2x2 optical fiber switch is used to control the flow of photons.
- Feed-Forward Control: When a herald photon is detected, a signal is sent (via FPGA logic) to the optical switch. This signal tells the switch to route the corresponding "idler" photon (the one intended for delivery) into the delay loop.
- Temporal Stacking: The switch is then reconfigured to keep the photon circulating in the loop. This process is repeated for subsequent laser pulses. If a photon is detected in a herald time bin, its twin is sent into the loop.
- Re-timing into a Single Bin: Crucially, the switch is timed to release the photons from the loop into the output path only at specific, synchronized intervals. In this experiment, they multiplexed over four consecutive pump pulses, meaning photons from four different time bins were effectively "stacked" and delivered into a single output time bin. This effectively increases the probability of having a photon in that specific output bin.
This approach is "resource-efficient" because it uses only one source and one switch, recycling the photons through the same delay loop.
Datasets
Not discussed in the paper. The experiments rely on the performance of their custom-built heralded single photon source and the optical components used in the multiplexing setup.
Key Insights and Ideas
- Temporal Stacking is Efficient: The most significant insight is that temporal multiplexing, by re-timing photons from a single source into a single output bin, can significantly boost the per-bin delivery probability. This bypasses the fundamental probabilistic limit of single-photon generation per pulse.
- Minimizing Resource Overhead: The use of a single switch and a simple fiber loop is a major advantage. It drastically reduces the complexity and cost compared to schemes requiring multiple sources or complex spectral/spatial routing.
- Trade-off Between Rate and Probability: While the per-bin probability increases, the overall rate of photons delivered per second decreases because the output bins are spaced further apart (e.g., every fourth time bin). The paper argues this is a worthwhile trade-off for experiments needing many synchronized photons.
- Loss is the Bottleneck: The performance of the multiplexing scheme is highly sensitive to the loss in the optical switch and the fiber loop. Each pass through the loop incurs loss, limiting the number of time bins that can be effectively multiplexed.
Key Results
- Enhanced Per-Bin Probability: The experiment demonstrated an enhancement in the per-bin single-photon delivery probability. With a switch loss of 1 dB and multiplexing over four time bins, they achieved up to a factor of 1.54(1) improvement in the probability of detecting a photon in the target output bin.
- Improved Heralding Efficiency: The heralding efficiency (a measure of how well the herald signal predicts the presence of a photon) also increased with higher heralding count rates when using the multiplexing scheme.
- Significant Advantage for Multi-Source Systems: While the per-bin improvement from a single source is modest, the paper highlights its dramatic impact when considering systems requiring many synchronized sources (N). For N=10 sources, their multiplexing scheme offers orders of magnitude improvement in the overall rate of delivering N photons simultaneously, with a break-even point at N=10 sources even with their current loss.
- Reduced Cross-Correlation: The re-timing process also improved the cross-correlation between the herald and idler photons, indicating better temporal alignment.
Limitations and Future Work
- Switch and Fiber Loss: The primary limitation is the insertion loss of the optical switch (1 dB) and the cumulative loss from multiple passes through the delay loop. This loss limits the number of time bins that can be effectively multiplexed.
- Reduced Output Rate: As mentioned, the temporal multiplexing reduces the overall photon delivery rate per second.
- Future Work: The authors suggest that advancements in low-loss fiber-integrated optical switches are crucial. Reducing switch loss would enable combining a much larger number of time bins, leading to even greater enhancements in per-bin probability and significantly improving the performance for multi-source quantum systems. They also note that optimizing the number of multiplexed bins (m) for a given number of sources (N) is an important consideration for future experiments.
Summary updated