Friday 22/5/2015, 11:00, room A5-01
Speaker: Valery Shchesnovich (UFABC)
Title: Partial distinguishability theory for photons and the Boson-Sampling computer
Multi-photon experiments in unitary linear networks necessitate systematic consideration of the effect of partial distinguishability of photons on probability distribution at network output. In this talk I introduce the partial distinguishability matrix for N photons in arbitrary (mixed) internal state at input of an unitary linear network and discuss its properties. I argue that the purity of the partial distinguishability matrix is a natural measure of quantum coherence of photons in a linear network, generalizing Mandel’s parameter for more than two photons. An analytical result for quantum coherence of single photons with fluctuating arrival times is presented. Another important application is the Boson Sampling computer with Linear Optics, where it is important to know how the computational complexity degrades with partial distinguishability of photons. I derive an upper bound on the trace distance between the output probability distribution of a realistic device with only partially indistinguishable bosons and the ideal Boson Sampling computer. Moreover, the lower bound follows from a physically clear conjecture on distinguishability of inputs with orthogonal partial distinguishability matrices. Finally, I present a scheme implementing the Scattershot version of the Boson Sampling computer with Fermionic Linear Optics and non-absorbing on-off type particle counting measurement.