Jason Saied, Jeffrey Marshall, Namit Anand, Eleanor G. Rieffel (Apr 23 2024).
Abstract: Highly pure and indistinguishable photons are a prerequisite for use in quantum information processing. We introduce protocols for the distillation of indistinguishable photons that offer a significant improvement over previous work, reducing distinguishability error rates by a factor of $n$, with resource requirements scaling linearly in $n$. We present the protocols, based on the discrete Fourier transform and Hadamard (Sylvester) matrices, then give both analytical and numerical results regarding their performance. We observe that the same symmetry properties governing suppression laws are instrumental in understanding the behavior of these distillation protocols. We also prove, adapting a result from the Hadamard case, that for the $n$-photon discrete Fourier transform with $n$ a prime power, the suppression laws are exactly characterized by the well-known Zero Transmission Law based on permutation symmetry.