Ancilla-Assisted Quantum Process Tomography

Abstract

Complete and precise characterization of a quantum dynamical process can be achieved via the method of quantum process tomography. Using a source of correlated photons, we have implemented several methods, each investigating a wide range of processes, e.g., unitary, decohering, and polarizing. One of these methods, ancilla-assisted process tomography (AAPT), makes use of an additional "ancilla system," and we have theoretically determined the conditions when AAPT is possible. Surprisingly, entanglement is not required. We present data obtained using both separable and entangled input states. The use of entanglement yields superior results, however.

Characterizing quantum gates via randomized benchmarking

“…One can estimate p directly using any of standard process tomography [15], ancilla-assisted/entanglement-assisted process tomography [48] or Monte-Carlo methods [23,24]. The tomography based schemes suffer from the unrealistic assumptions of negligible state-preparation and measurement errors, and clean ancillary states/operations.…”

Quantum-process tomography: Resource analysis of different strategies

“…This means that entanglement is not a necessary property of the input state ρ in AAPT. Indeed, many of the viable input states are not entangled, such as Werner states [9]. However, it has been argued and also experimentally verified that use of maximally entangled pure states offers the best performance.…”

“…Generally, within the AAPT scheme, we attach an auxiliary system (ancilla), B, to our principal system, A, and prepare the combined system in a single state such that complete information about the dynamics can be imprinted on the final state [8,9]. Then by performing quantum state tomography in the extended Hilbert space of H AB , one can extract complete information about the unknown map acting on the principal system.…”

“…Indeed, it is easy to show that a state ρ can be used as input for AAPT iff ρ has maximal Schmidt number, i.e. Sch(ρ) = d 2 [9]. 2 The faithfulness condition 1 Choosing {Em} = {|i j|} results in: χ = dρ E [33].…”

“…A further unavoidable step in indirect methods is the application of an inversion map on the final output data. Standard Quantum Process Tomography (SQPT) [1,5,6] and Ancilla-Assisted Process Tomography (AAPT) [7,8,9,10] belong to this class. On the other hand, in direct methods each experimental outcome directly provides information about properties of the underlying dynamics, without the need for state tomography.…”