| name | decoded-quantum-interferometry-bounded-degree |
| description | Complexity-theoretic benchmark for decoded quantum interferometry (DQI), QAOA, and classical heuristics on bounded-degree max-Ek-LINSAT. Extends NP-hardness to arbitrary finite fields F_q with bounded degree D, proving hardness to exceed r/q + O(1/sqrt(D)). Identifies 1/sqrt(D log D) classical decoder barrier vs. 1/sqrt(D) quantum decoder scaling. arXiv: 2606.13570 |
| category | quantum/optimization |
| metadata | {"arxiv_id":"2606.13570","authors":"Maximilian J. Kramer, Carsten Schubert, Jens Eisert","subjects":"quant-ph,cs.CC,cs.DS","published_date":"2026-06-11"} |
Context
For general max-k-XORSAT with k>=3, no polynomial-time algorithm can substantially beat random guessing on worst-case instances unless P=NP. However, bounded-degree instances (each variable appears in at most D constraints) are algorithmically easier — polynomial-time algorithms can beat the random baseline by O(1/sqrt(D)). This work extends the hardness analysis to bounded-degree max-Ek-LINSAT over arbitrary finite fields F_q and establishes the complexity-theoretic benchmark for DQI, QAOA, and classical heuristics.
Core Methodology
Hardness Extension to Finite Fields
For max-Ek-LINSAT(q,r) with bounded degree D over F_q:
- It is NP-hard to exceed r/q + O_{q,r}(1/sqrt(D))
- This extends the Boolean case (Trevisan/Barak) to arbitrary finite fields
- Provides the complexity-theoretic ceiling for any algorithm (classical or quantum)
DQI Decoder Scaling Analysis
-
Classical Decoder Barrier: DQI with classical decoders faces an information-theoretic 1/sqrt(D log D) barrier
- This prevents matching the complexity-theoretic
1/sqrt(D) scaling
- Represents a fundamental limitation of classical post-processing
-
Quantum Decoder Compatibility: DQI with quantum decoders is compatible with the 1/sqrt(D) scaling
- Identifies quantum decoding as the key ingredient for matching complexity-theoretic bounds
- Provides a clear target for quantum advantage demonstrations
Implications for QAOA
- QAOA on bounded-degree instances is similarly constrained by the
1/sqrt(D) ceiling
- Any quantum advantage is confined to improving the constant prefactor
- The hardness result establishes when QAOA can/cannot outperform classical methods
Algorithmic Benchmarking Framework
For bounded-degree optimization:
1. Compute the random assignment baseline: r/q
2. The hardness ceiling: r/q + O(1/sqrt(D))
3. Classical algorithm target: r/q + O(1/sqrt(D)) — achievable
4. DQI classical decoder: r/q + O(1/sqrt(D log D)) — suboptimal
5. DQI quantum decoder: r/q + O(1/sqrt(D)) — optimal
6. QAOA: depends on depth and angle optimization
Implementation Steps
Step 1: Problem Formulation
Input: Linear system over F_q with k variables per equation, max degree D
Output: Assignment maximizing satisfied equations
Parameters:
- q: Field size (q=2 for Boolean)
- k: Number of variables per equation
- D: Maximum degree (variable occurrences)
- r: Target satisfaction ratio
Step 2: Classical Baseline Computation
- Random assignment achieves r/q expected satisfaction
- For k>=3, NP-hard to beat this by more than O(1/sqrt(D)) on worst-case instances
- Polynomial-time algorithms can achieve O(1/sqrt(D)) improvement on bounded-degree instances
Step 3: DQI Protocol
DQI Protocol:
1. Encode the optimization problem into a quantum circuit
2. Apply the quantum interference pattern
3. Decode measurement outcomes:
a. Classical decoder: achieves O(1/sqrt(D log D)) — below optimal
b. Quantum decoder: achieves O(1/sqrt(D)) — matches hardness ceiling
4. Return best assignment
Step 4: QAOA Comparison
QAOA Protocol:
1. Construct cost Hamiltonian from max-Ek-LINSAT instance
2. Optimize variational angles (p layers)
3. Sample and decode
4. Performance bounded by O(1/sqrt(D)) ceiling
Step 5: Benchmarking
Benchmark Protocol:
1. Generate (k,D)-regular instances over F_q
2. Run: random baseline, classical algorithm, DQI (classical/quantum decoder), QAOA
3. Compare achieved approximation ratios against hardness ceiling
4. Identify regimes where quantum methods show advantage
Pitfalls
- Bounded-Degree Assumption: Results only apply to bounded-degree instances. For unbounded-degree instances, the hardness is much stronger (NP-hard to beat 1/2 by any constant for k>=3). Fix: Verify degree bounds before applying these results.
- Constant Prefactor Focus: On bounded-degree instances, quantum advantage is only in the constant prefactor, not asymptotic scaling. Implication: Advantage may be small and hard to demonstrate experimentally.
- Classical Decoder Barrier: The
1/sqrt(D log D) barrier for classical decoders is information-theoretic, not computational. Implication: No classical decoding strategy (regardless of computational power) can overcome this barrier within the DQI framework.
- Field Generalization: The extension to arbitrary F_q introduces q,r-dependent constants. Implication: The exact scaling constants depend on field parameters.
- Worst-Case vs. Average-Case: Hardness results are worst-case. Implication: Average-case instances may be easier. Distinguish between worst-case hardness and practical performance.
Verification
- Hardness Bound: Verify the NP-hardness proof for max-Ek-LINSAT(q,r) with bounded degree D.
- DQI Classical Barrier: Confirm the
1/sqrt(D log D) information-theoretic barrier for classical decoders.
- DQI Quantum Compatibility: Verify that quantum decoders can achieve
1/sqrt(D) scaling.
- QAOA Ceiling: Confirm QAOA is similarly bounded by
1/sqrt(D).
- Empirical Validation: Test on concrete bounded-degree instances and compare achieved ratios against bounds.
Activation
decoded quantum interferometry, bounded degree optimization, max-LINSAT hardness, quantum decoder advantage, DQI classical barrier, QAOA bounded degree, finite field satisfiability, approximation hardness quantum, complexity-theoretic quantum benchmark, information-theoretic decoder barrier