Source-linked AI summary

Fixed-point quantum search with an optimal number of queries

Theodore J. Yoder, Guang Hao Low, Isaac L. Chuang

arXiv:1409.3305v2quant-ph

TL;DR

Grover search offers a quadratic speedup, but its iteration count depends on the unknown target-state fraction and can overcook the state. The paper constructs fixed-point amplitude amplification with tunable bounded error, while retaining optimal query scaling and avoiding overcooking.

  • Problem

    Grover’s search can overcook the state when the number of marked items is unknown, while existing fixed-point algorithms lose the quadratic speedup.

  • Method

    The authors use the polynomial method to tune phases in generalized Grover reflections, constructing sequences with adjustable error δ and fixed-point behavior.

  • Results

    The fixed-point search cannot be overcooked and achieves optimal time scaling, providing a quadratic advantage over classical unordered search.

  • Takeaways & Limitations

    The algorithm can serve as a subroutine for amplitude amplification tasks without estimating the correct iteration count or remaking the initial state.

  • Takeaways & Limitations

    The π/3-algorithm special case at δ = 0 has classical query scaling, so the stated optimality does not extend to that case.

Abstract

from arXiv · show

Grover's quantum search and its generalization, quantum amplitude amplification, provide quadratic advantage over classical algorithms for a diverse set of tasks, but are tricky to use without knowing beforehand what fraction $λ$ of the initial state is comprised of the target states. In contrast, fixed-point search algorithms need only a reliable lower bound on this fraction, but, as a consequence, lose the very quadratic advantage that makes Grover's algorithm so appealing. Here we provide the first version of amplitude amplification that achieves fixed-point behavior without sacrificing the quantum speedup. Our result incorporates an adjustable bound on the failure probability, and, for a given number of oracle queries, guarantees that this bound is satisfied over the broadest possible range of $λ$.

Loading 1409.3305v2…