Bosonsampling with controllable distinguishability of photons

The BosonSampling Problem is to sample output photon-coincidence probabilities given vacuum and single-photon inputs to a passive interferometer with more channels than photons. This problem is classically hard to simulate as these probabilities are weighted by computationally hard permanents of sub-matrices of the interferometer transition matrix yet efficient to execute quantumly. Our innovation [1,2] introduces distinguishability between photons by controlling arrival times of otherwise identical photons in order to test the model, assess sampling errors and generalize BosonSampling beyond permanent-weighted to immanant-weighted probabilities.