Analysis of circuit imperfections in BosonSampling

Abstract

BosonSampling is a problem where a quantum computer offers a provable speedup over classical computers. Its main feature is that it can be solved with current linear optics technology, without the need for a full quantum computer. In this work, we investigate whether an experimentally realistic BosonSampler can really solve BosonSampling without any fault-tolerance mechanism. More precisely, we study how the unavoidable errors linked to an imperfect calibration of the optical elements affect the final result of the computation. We show that the fidelity of each optical element must be at least $1 - O(1/n^2)$, where $n$ refers to the number of single photons in the scheme. Such a requirement seems to be achievable with state-of-the-art equipment.