Figure 2

(a) Experimentally measured time-gated detector events for Charlie. The time-gated signal count rate is the detector click rate summed over all four of Charlie’s signal SPADs, filtered by a window of ± 1 ns around the expected pulse arrival time. The “loss corrected gated rate” is the calculated time-gated count rate at the signal output of the multiport. (b) Rates of successful and failed measurements for one recipient. “USE success” means that at least one state was correctly excluded using quantum state elimination. That is, the state Alice actually sent was not excluded, and at least one other state was excluded. “USE failure” means that the state Alice actually sent was excluded. “USD success” is the rate of successful unambiguous state discrimination (the correct state was obtained) while “USD failure” is the rate of unsuccessful unambiguous state discrimination (the obtained state differs from that sent by Alice). Data points represent experimental results and dashed lines are theoretical predictions [18]. Experimental data are averaged over several measurements and error bars in the count rate are the standard deviation. Horizontal error bars for the mean photon number are dominated by a worst-case assumption that the pulse-to-pulse variation in the output power of our laser is the experimentally measured maximum of ± 1.5 % .