Tutorial 3 — Routing Strategy Comparison

Objective: Run the same topology under three routing strategies to see how strategy choice affects latency vs fidelity trade-offs.

qnet-core supports three routing strategies that prioritize different aspects of network performance. Each strategy makes different trade-offs between speed and quality of the established entanglement.

Define an Asymmetric Network

from qnet_core import QNetEngine, NodeDefinition, LinkDefinition, StrategyType

engine = QNetEngine()

nodes = [
    NodeDefinition(id="Alice", memory_lifetime_t2=1.0),
    NodeDefinition(id="R1",   memory_lifetime_t2=0.8),
    NodeDefinition(id="R2",   memory_lifetime_t2=0.6),
    NodeDefinition(id="Bob",  memory_lifetime_t2=1.0),
]

links = [
    # Short but noisy link (high fidelity, fast generation)
    LinkDefinition("Alice", "R1", distance_km=5.0, base_fidelity=0.97, generation_rate_hz=2_000.0),
    # Long but clean link (lower fidelity, slower generation)
    LinkDefinition("R1", "R2", distance_km=30.0, base_fidelity=0.80, generation_rate_hz=200.0),
    # Medium link
    LinkDefinition("R2", "Bob", distance_km=15.0, base_fidelity=0.90, generation_rate_hz=800.0),
]

engine.define_network(nodes, links)

Run Under Each Strategy

strategies = [
    StrategyType.LowestLatency,
    StrategyType.HighestFidelity,
    StrategyType.HighestSuccess,
]

for strategy in strategies:
    stats = engine.simulate(
        from_node="Alice",
        to="Bob",
        fidelity_target=0.75,
        max_latency_ms=10_000.0,
        runs=500,
        strategy=strategy,
        seed=42,
    )

    print(f"{strategy:<20} {stats.empirical_success_rate:>7.1%} "
          f"{stats.mean_latency_ms:>9.1f} ms {stats.mean_fidelity:>9.4f}")

Strategy Reference

Strategy	Priority	How It Works
LowestLatency	Speed	Minimizes total hop distance; fastest route even with lower per-link fidelity
HighestFidelity	Quality	Maximizes end-to-end fidelity via purification; may take longer but ensures quality
HighestSuccess	Reliability	Prefers links/paths with higher generation success rates (stochastic optimization)

Typical Results

On an asymmetric network like the one above, you'll typically see:

• LowestLatency: Shortest total path, lowest latency, but possibly lower final fidelity
• HighestFidelity: May take a longer route with more purification hops; highest fidelity at cost of latency
• HighestSuccess: Balances both by preferring high-rate links with good fidelity — often the best all-around choice

What's next

• Monte Carlo Ensemble — get statistics over many runs for any strategy
• Built-in Topology Generators — use pre-built topologies for faster setup