Title: Experimental network advantage for quantum conference key agreement

Abstract: One of the great promises of quantum technology is the development of quantum networks, which will allow global distribution of entanglement for tasks such as distributed quantum computing, distributed quantum sensing and quantum-secure communication. To leverage the full potential of quantum networks we require protocols that draw an efficiency advantage from genuine multi-partite entanglement as opposed to strictly pair-wise correlations such as Bell states. Multi-user entanglement such as Greenberger-Horne-Zeilinger (GHZ) states have already found application in quantum conference key agreement, quantum secret sharing and quantum communication complexity problems. However, a true network advantage has not yet been achieved. In this work we create a six-photon graph-state network from which we derive either a four-user GHZ state for direct quantum conference key agreement or the required amount of Bell pairs for the equivalent pair-wise protocol. We show that the GHZ-state protocol has a more than two-fold rate advantage by only consuming half the amount of network resources per secure conference key bit.