Quantum property testing for bounded-degree graphs

We study quantum algorithms for testing bipartiteness and expansion of bounded-degree graphs. We give quantum algorithms that solve these problems in time O(N^(1/3)), beating the Omega(sqrt(N)) classical lower bound. For testing expansion, we also prove an Omega(N^(1/4)) quantum query lower bound, thus ruling out the possibility of an exponential quantum speedup. Our quantum algorithms follow from a combination of classical property testing techniques due to Goldreich and Ron, derandomization, and the quantum algorithm for element distinctness. The quantum lower bound is obtained by the polynomial method, using novel algebraic techniques and combinatorial analysis to accommodate the graph structure.

💡 Research Summary

The paper investigates quantum algorithms for two fundamental property‑testing problems on bounded‑degree graphs: bipartiteness and expansion. In the classical setting, Goldreich and Ron showed that testing these properties requires Θ(√N) queries, and a matching Ω(√N) lower bound is known. The authors ask whether quantum computation can break this barrier.

For bipartiteness, they adapt the classical random‑walk based tester but replace the classical collision check with the quantum element‑distinctness algorithm. The tester randomly selects a vertex, performs a short random walk of length ℓ = Θ(N¹ᐟ³), and records the labels of visited vertices in quantum superposition. Detecting whether any label repeats is exactly the element‑distinctness problem, which can be solved with O(N¹ᐟ³) quantum queries. By carefully derandomizing the walk selection and ensuring that the walk explores a sufficiently large portion of the graph, they prove that the algorithm accepts bipartite graphs with high probability and rejects graphs that are ε‑far from bipartite with comparable probability. Consequently, bipartiteness can be tested in O(N¹ᐟ³) time, a cubic‑root speed‑up over the optimal classical bound.

The expansion tester follows a similar blueprint. Classical expansion testers typically rely on spectral methods or on counting the number of distinct vertices reached by short random walks. The quantum algorithm again performs short random walks from randomly chosen seeds, but it uses the element‑distinctness subroutine to verify that the set of visited vertices contains few collisions. If the graph is a good expander, short walks rapidly spread and collisions are unlikely; if the graph is far from any expander, many collisions appear. The same O(N¹ᐟ³) query complexity is achieved, providing a quantum advantage for expansion testing as well.

Beyond algorithmic upper bounds, the paper establishes a non‑trivial quantum lower bound for expansion testing. Using the polynomial method, the authors construct a family of Boolean functions that encode the adjacency structure of bounded‑degree graphs. They show that any quantum algorithm that distinguishes an ε‑expander from a graph that is ε‑far from any expander must compute a polynomial of degree at least Ω(N¹ᐟ⁴). This lower bound is derived by a novel algebraic framework that captures the combinatorial constraints imposed by the bounded degree, extending previous lower‑bound techniques that were limited to unstructured collision problems. The result rules out the possibility of an exponential quantum speed‑up for expansion testing, placing the achievable quantum advantage firmly between N¹ᐟ⁴ and N¹ᐟ³.

Technically, the paper makes three key contributions: (1) it shows how to embed classical property‑testing procedures into quantum circuits by replacing collision detection with the optimal quantum element‑distinctness algorithm; (2) it introduces a derandomization technique that reduces the amount of randomness required while preserving the statistical guarantees needed for property testing; and (3) it develops a new polynomial‑method lower‑bound argument that respects the graph’s structural constraints, thereby establishing the Ω(N¹ᐟ⁴) quantum query lower bound for expansion.

The authors also discuss the broader implications of their work. The O(N¹ᐟ³) quantum testers demonstrate that quantum computers can provide polynomial‑speedups for natural graph‑property problems, even when the input is given in the adjacency‑list model. The Ω(N¹ᐟ⁴) lower bound, however, indicates that quantum algorithms cannot achieve exponential improvements for these tasks, suggesting a nuanced landscape where quantum advantage is problem‑dependent. Future directions include extending the techniques to other graph properties such as k‑colorability, cycle detection, or subgraph isomorphism, and exploring whether similar quantum‑classical separations exist in the dense‑graph model or for property testing with two‑sided error.

Overall, the paper delivers a comprehensive study of quantum property testing on bounded‑degree graphs, providing both concrete algorithmic improvements and rigorous limitations, and thereby advancing our understanding of the true power of quantum computation in the realm of sublinear‑time graph algorithms.