⚛️ quantum physics

Fidelity-Age-Aware Scheduling in Quantum Repeater Networks

This paper introduces the "Fidelity-Age" (FA) metric to measure the freshness of high-fidelity entanglement in quantum repeater networks and proposes two lightweight scheduling algorithms that minimize entanglement age while maintaining throughput and resource constraints.

Original authors: Ozgur Ercetin, Zafer Gedik

Published 2026-02-11

📖 4 min read🧠 Deep dive

CC BY 4.0

Original authors: Ozgur Ercetin, Zafer Gedik

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are running a global pizza delivery service, but there’s a catch: the pizzas are made of ice.

If you deliver a pizza too late, it doesn't just get cold; it melts into a puddle, making it completely useless to the customer. In the world of quantum computing, scientists are trying to send "entangled" particles (the pizzas) across long distances. These particles are incredibly fragile. If they sit in a "memory" (a delivery bag) for too long, they lose their "fidelity" (their shape/quality) and melt away.

This paper, written by researchers at Sabancı University, introduces a new way to manage these "melting" deliveries.

1. The Problem: The "Melting Pizza" Dilemma

Currently, most people managing quantum networks focus on Throughput—which is like trying to deliver as many pizzas as possible per hour.

The problem is that if you only focus on speed, you might spend all your time delivering pizzas to the houses right next door because it’s "easy" and "fast." Meanwhile, the customer living on the far side of town is ignored. Their pizza sits in the bag longer and longer until it eventually melts. In quantum terms, this is called "starvation" or "extreme age." The customer gets nothing, and the network becomes unfair and unstable.

2. The Innovation: The "Fidelity-Age" (FA) Metric

The researchers created a new ruler to measure success called Fidelity-Age (FA).

Instead of just asking, "How many pizzas did we deliver?" they ask, "How long has it been since we delivered a perfect, non-melted pizza to this specific customer?"

This metric combines two things:

Fidelity: Is the pizza still a pizza, or is it a puddle? (Is the quantum state still usable?)
Age: How long has the customer been waiting?

By tracking this, the network doesn't just care about quantity; it cares about freshness and reliability.

3. The Solution: The "Smart Dispatcher" (The Schedulers)

The researchers tested two new "Smart Dispatchers" (algorithms) to manage the deliveries:

The "Threshold" Dispatcher (FA-THR): This dispatcher has a rule: "If a customer has been waiting longer than 10 minutes, they become a priority, no matter what." It prevents anyone from being ignored for too long.
The "Index" Dispatcher (FA-INDEX): This is a more sophisticated dispatcher. It looks at every order and gives it a "priority score" based on how likely the delivery is to succeed multiplied by how hungry (old) the customer is. It’s like a math equation that balances "How easy is this delivery?" with "How much does this person need us right now?"

4. The Results: Why It Matters

When they ran simulations (digital "stress tests"), the results were impressive:

No More "Puddles": The old way (focusing only on speed) led to some customers waiting so long their "pizzas" were essentially gone. The new way kept the "age" of the deliveries very low and stable.
Efficiency without Chaos: Usually, if you try to be "fair," you slow down your total number of deliveries. But these new dispatchers were "Pareto-efficient"—meaning they kept the delivery speed high while also making sure no one was left waiting in the dark.
Resilience: Even when the "delivery bags" (quantum memory) were poor quality or the "roads" (fiber optic cables) were long and difficult, the smart dispatchers adjusted their routes to keep the network running smoothly.

The Big Picture

In the future, when we have a "Quantum Internet" that connects supercomputers across the globe, we can't just rely on old-school methods. We need a system that understands that timing is everything. This paper provides the mathematical blueprint for a network that is not just fast, but also reliable, fair, and—most importantly—always fresh.

Technical Summary: Fidelity-Age–Aware Scheduling in Quantum Repeater Networks

1. Problem Statement

Quantum repeater networks are designed to distribute entanglement over long distances to support applications like distributed quantum computing and high-rate QKD. However, unlike classical data packets, entangled states are transient physical resources subject to decoherence (fidelity decay over time) and stochasticity (probabilistic success of entanglement swapping and purification).

Existing research primarily focuses on optimizing throughput (the rate of entanglement generation) or end-to-end fidelity. The authors identify a critical gap: the lack of a metric to quantify the "freshness" of entanglement. A network might maintain high throughput, but if the deliveries are highly irregular, certain users may experience extreme delays, receiving "stale" entanglement that is barely usable. The core problem is how to schedule limited quantum resources (memory and transmission attempts) to ensure timely, high-fidelity entanglement delivery while preventing "starvation" of specific nodes.

2. Methodology

The paper introduces a novel framework based on the following components:

The Fidelity-Age (FA) Metric: Inspired by the classical "Age of Information" (AoI), the FA metric measures the time elapsed since the last successful delivery of an entangled pair whose fidelity $F$ meets or exceeds a minimum threshold $F_{min}$ . This metric uniquely couples temporal freshness with physical-layer quality.
Mathematical Modeling:
- Renewal Theory: The authors model usable entanglement deliveries as "renewal events." They derive a renewal identity that mathematically links the slot-level success probability ( $p_{succ}$ ) to the long-run average FA.
- Stochastic Control: The scheduling problem is formulated as a stochastic control problem aimed at minimizing the weighted sum of average FA across all requests, subject to constraints on attempt budgets ( $R$ ), local resource exclusivity, and quantum memory limits.
Proposed Schedulers: To solve the control problem without the computational overhead of global optimization, the authors propose two lightweight, Lyapunov-drift-inspired heuristics:
1. FA-THR ( $\tau$ ): A threshold-based policy that prioritizes flows whose age exceeds a certain value $\tau$ .
2. FA-INDEX ( $\beta$ ): A prioritization policy that ranks flows by a weighted index: $\text{Utility} / (1 + \beta \cdot \text{Age})$ . This smoothly balances instantaneous throughput against the need to refresh stale flows.

3. Key Contributions

New Metric Definition: The formal introduction of the Fidelity-Age (FA) metric, providing a physically grounded way to measure entanglement usability.
Analytical Foundation: The derivation of the FA renewal identity, which allows for the prediction of long-term performance based on short-term success probabilities.
Control Framework: A formalization of the scheduling problem as a stochastic optimization task under realistic quantum hardware constraints (decoherence, finite memory, and probabilistic BSMs).
Practical Algorithms: The development of the FA-THR and FA-INDEX schedulers, which are computationally efficient ( $O(|P| \log |P|)$ ) and suitable for real-time SDN (Software-Defined Network) controllers.

4. Results

The authors conducted extensive Monte Carlo simulations on a $3 \times 3$ repeater grid. Key findings include:

Mitigation of Extreme Latency: While conventional greedy policies (TP-MAX, FID-MAX) maximize throughput, they suffer from "heavy-tailed" age distributions and high starvation rates (up to 25%). In contrast, FA-aware schedulers reduce extreme-age events (CVaR95) by up to two orders of magnitude.
Pareto Efficiency: The FA-aware policies occupy the Pareto frontier, meaning they achieve throughput levels comparable to greedy policies while significantly improving temporal stability and fairness.
Resilience to Decoherence: In regimes with short coherence times ( $T_2$ ), FA-aware schedulers outperform others by prioritizing the "refreshing" of links before they decohere beyond the $F_{min}$ threshold.
Fairness: FA-INDEX achieved a high Jain’s Fairness Index ( $\approx 0.91$ ), ensuring that even under high network load, no single request is starved of entanglement.

5. Significance

This work shifts the paradigm of quantum network management from purely rate-centric (how much entanglement?) to freshness-centric (how timely and reliable is the entanglement?). By providing a tractable mathematical link between physical decoherence and network-layer scheduling, the paper establishes a foundation for building reliable, stable, and fair large-scale quantum internet architectures. The FA metric and its associated schedulers offer a scalable way to manage the inherent volatility of quantum physical layers.