Thread by @fgrosshans

Now at #IWQNS, Don Towsley from @UMassAmherst on Entanglement routing and switching in quantum networks #LTQI

@UMassAmherst Don Towsley looks at repeaters/routers/switchs
With m links/link
Phase 1: link entanglement succeeds with p=1 – (1 — p₀)^m
Phase 2: splicing the links succeeds with probability q
#LTQI #IWQNS

@UMassAmherst Don Towsley study this in a grid network, with a single mode per link, one memory/mode, and a probability p to generate entanglement. #LTQI #IWQNS

@UMassAmherst Don Towsley Compute the rate dependence on p and q, with a simple greedy shortest path algorithm. It requires global information. We see percolation effect for q=1 around p∼.5.
#IWQNS #LTQI

@UMassAmherst Don Towsley: With imperfect Bell measurement (q<1). We have a trivial upper bound and an easy algorithm, but there is a large gap.
#LTQI #IWQNS

@UMassAmherst Don Towsley looks at a simple local entanglement flow rooting, based on potential shorted path connection (based on Euclidean distance to Alice and Bob) plus any other path if possible. It gives results better than shortest possible path.
But exp decay with distance
#IWQNS #LTQI

@UMassAmherst Don Towsley now looks at multiflow routing. One can optimize the protocol for this, the best is through spatial division (attribution of some switches to Alice1, Alice2, Bob1 or Bob2)
#LTQI #IWQNS

Don Towsley now looks at a switch with may users, with a FIFO-like policy. With a model with a simple birth-death process. The expected number of stored qubit is quite small (∼½) for an essentially infinite capacity.
#LTQI #IWQNS

Don Towsley also looked at heteregenous links. Mismatch of links leads to storage of entanglement for the fast links.
He also looks at decoherence with a very simple model, and tripartite entanglement switching.
#LTQI #IWQNS

@threadreaderapp Please kindly unroll