[Paper Review] Multidimensional Quantum Walks, with Application to $k$-Distinctness

Stacey Jeffery, Sebastian Zur|arXiv (Cornell University)|Aug 29, 2022

Quantum Computing Algorithms and Architecture4 citations

TL;DR

This paper introduces a novel multidimensional quantum walk framework that achieves time complexity matching the best-known query complexity for $k$-distinctness, $\widetilde{O}(n^{\frac{3}{4}-\frac{1}{4}\frac{1}{2^{k}-1}})$, by extending the electric network framework. It resolves a decade-old open problem by constructing an efficient time-optimal quantum algorithm for $k$-distinctness and demonstrates exponential speedup on the welded trees problem.

ABSTRACT

While the quantum query complexity of $k$-distinctness is known to be $O\left(n^{3/4-1/4(2^k-1)} ight)$ for any constant $k \geq 4$, the best previous upper bound on the time complexity was $\widetilde{O}\left(n^{1-1/k} ight)$. We give a new upper bound of $\widetilde{O}\left(n^{3/4-1/4(2^k-1)} ight)$ on the time complexity, matching the query complexity up to polylogarithmic factors. In order to achieve this upper bound, we give a new technique for designing quantum walk search algorithms, which is an extension of the electric network framework. We also show how to solve the welded trees problem in $O(n)$ queries and $O(n^2)$ time using this new technique, showing that the new quantum walk framework can achieve exponential speedups.

Motivation & Objective

• To close the gap between query complexity and time complexity for the $k$-distinctness problem, which had remained open for ten years despite optimal query bounds.
• To develop a new quantum walk framework capable of achieving exponential speedups, unlike prior frameworks limited to quadratic speedups.
• To provide a constructive, time-efficient quantum algorithm for $k$-distinctness matching Belovs' query complexity upper bound.
• To demonstrate the power of the new framework by solving the welded trees problem in $O(n)$ queries and $O(n^2)$ time, achieving exponential speedup.

Proposed method

• Proposes a multidimensional quantum walk framework as an extension of the electric network framework, enabling more complex and efficient quantum walk constructions.
• Uses phase estimation and structured graph decomposition to simulate quantum walks across multiple dimensions, with time and error bounds carefully controlled.
• Employs a layered graph structure $G$ with $k$ levels, where each level $\ell$ corresponds to a stage of candidate $k$-tuples, and defines transition maps between levels.
• Introduces a transition subroutine with controlled error $\epsilon_{u,v} = O(n^{-\kappa})$ for $\kappa > 2$, ensuring negligible error accumulation.
• Applies Theorem 3.10 to bound the total time complexity using weighted walk cost $\mathcal{W}^T$ and resistance $\mathcal{R}^T$, achieving $\widetilde{O}(n^{\frac{3}{4}-\frac{1}{4}\frac{1}{2^{k}-1}})$ time.
• Optimizes parameters $t_\ell$ across levels to balance query and time costs, with $t_\ell = n^{\frac{3}{4}-\frac{1}{4}\frac{1}{2^{k}-1}-\sum_{\ell'=2}^{\ell}\frac{2^{k-1-\ell'}}{2^{k}-1}}$.

Experimental results

Research questions

• RQ1 Can the time complexity of $k$-distinctness be matched to its best-known query complexity of $O(n^{\frac{3}{4}-\frac{1}{4}\frac{1}{2^{k}-1}})$?
• RQ2 Can a quantum walk framework be designed to achieve exponential speedups beyond the quadratic limit of prior frameworks?
• RQ3 Is it possible to construct a time-efficient quantum algorithm for $k$-distinctness that realizes Belovs' dual adversary-based query bound?
• RQ4 Can the new multidimensional quantum walk framework solve classically hard problems like the welded trees problem efficiently?

Key findings

• The time complexity of $k$-distinctness is bounded by $\widetilde{O}(n^{\frac{3}{4}-\frac{1}{4}\frac{1}{2^{k}-1}})$, matching the best-known query complexity up to polylogarithmic factors.
• The new multidimensional quantum walk framework enables exponential speedup on the welded trees problem, solving it in $O(n)$ queries and $O(n^2)$ time.
• The framework achieves this by extending the electric network framework to support multi-level, multi-dimensional walk structures with controlled error and resistance.
• The algorithm constructs a layered graph where each level corresponds to a stage of $k$-tuple candidate generation, with optimized parameters $t_\ell$ minimizing total time.
• The framework overcomes the quadratic speedup barrier of prior quantum walk models by allowing non-local, multi-dimensional transitions that capture complex dependencies.
• The construction is fully constructive, providing an explicit quantum circuit with efficient state preparation and transition subroutines, resolving the long-standing open problem of time efficiency for $k$-distinctness.