Entanglement swapping with independent sources over an optical-fiber network

Establishing entanglement between two remote systems by the method of entanglement swapping is an essential step for a long-distance quantum network. Here we report a field-test entanglement swapping experiment with two independent telecommunication band entangled photon-pair sources over an optical fiber network in Hefei. The two sources are located at two nodes that are $12.5\mathrm{km}$ apart and the Bell-state measurement is performed at a third location which is connected to the two source nodes with 14.7-km and 10.6-km optical fibers, respectively. The observed average visibility is $79.9\pm 4.8\%$, which is sufficient for the violation of Bell inequalities. Furthermore, with the swapped entanglement, we demonstrate a source-independent quantum key distribution, which is also immune to any detection attacks at the measurement site.

Figure 1

Bird's-eye view and schematic of entanglement swapping in the Hefei optical fiber network. The locations of Alice (Hefei Innovation Industrial Park), Bob (University of Science and Technology of China), and Eve (Hefei Software Park) are marked on the satellite image. Alice and Bob are separated by 12.5 km and connected to Eve using 14.7-km and 10.6-km deployed optical fibers, respectively. Both Alice and Bob prepare time-bin-entangled photon pairs, measure the idler photons, and send their twin signal photons to Eve for BSM. The synchronization signals (Sync) are generated by Eve and then distributed to Alice and Bob through the optical fibers. The single photons and strong laser pulses are transmitted in different optical fibers, represented by the yellow line and purple lines, respectively.

Figure 2

Fourfold coincidence count probabilities as a function of the phase of Alice's MZI. The error bars indicate one standard deviation calculated from measured counts assuming Poissonian detection statistics. Each data point is accumulated for more than 10 h. The visibilities of the fitted curve are $81.2\pm 6.2\%$ and $78.6\pm 7.3\%$ for measured results with ${\varphi }_B=0$ and ${\varphi }_B=\pi /2$, respectively.

Figure 3

Experimental results. (a) Sifted key in the time basis (T) and the energy basis (E) per hour, with the numbers of wrong (W) and correct (C) bits. (b) Secure key rate per sifted key with one B step of the entanglement swapping based QKD as well as the simulation result. The circle and star correspond to the experimental result in the energy basis (60 bits) and the time basis (58 bits), respectively. In the simulation, $N^e=N^t=2500$ and $e_b^e=e_b^t$.