Quantum State Discrimination Using the Minimum Average Number of Copies

TL;DR

Prior work minimized discrimination error for fixed resources, whereas this paper defines guaranteed bounded error discrimination to minimize average copies under a fixed error bound. It introduces and tests a globally optimal fixed measurement scheme, finding it generally matches or outperforms other strategies, with a practical advantage when errors are small.

• Problem

  Guaranteed bounded error discrimination minimizes the average number of copies needed to keep error below a preset bound, unlike prior minimum-error formulations with fixed resources.

  Source: p. 1

• Method

  The paper applies existing local non-adaptive schemes, derives a new globally optimal fixed measurement scheme, and evaluates it theoretically and experimentally for two pure nonorthogonal states.

  Source: pp. 1–3

• Results

  The new scheme performs at least as well as other local non-adaptive strategies, usually outperforms the minimum-error-optimal adaptive strategy, and has better scaling in the small-error regime.

  Source: pp. 1, 2, 5

• Takeaways & Limitations

  Resource minimization can reverse which discrimination strategy is preferable compared with error minimization, making task-specific measurement design important.

  Source: p. 5

  The study restricts the states to two specified nonorthogonal pure qubit states and focuses experimentally and numerically on equal prior probabilities.

  Source: p. 2