Efficient Pure State Quantum Tomography from Five Orthonormal Bases

TL;DR

This paper presents a fixed, efficient method for pure quantum state tomography using five orthonormal bases, which is robust to noise and does not depend on the unknown state.

Contribution

The authors explicitly construct five orthonormal bases enabling fixed, efficient pure state tomography, independent of the unknown state, and provide convex programs for reconstruction.

Findings

The measurement scheme is robust against noise.

The scheme allows for efficient reconstruction of pure states.

It is a fixed measurement scheme, unlike adaptive methods.

Abstract

For any finite dimensional Hilbert space, we construct explicitly five orthonormal bases such that the corresponding measurements allow for efficient tomography of an arbitrary pure quantum state. This means that such measurements can be used to distinguish an arbitrary pure state from any other state, pure or mixed, and the pure state can be reconstructed from the outcome distribution in a feasible way. The set of measurements we construct is independent of the unknown state, and therefore our results provide a fixed scheme for pure state tomography, as opposed to the adaptive (state dependent) scheme proposed by Goyeneche et al. in [Phys. Rev. Lett. 115, 090401 (2015)]. We show that our scheme is robust with respect to noise in the sense that any measurement scheme which approximates these measurements well enough is equally suitable for pure state tomography. Finally, we present two convex programs which can be used to reconstruct the unknown pure state from the measurement outcome distributions.