Figure 1

Experimental setup for two photon fault-tolerant quantum key distribution protocol. The 394-nm UV pulses are produced by frequency doubling the 788-nm pulses of the mode-locked laser using a nonlinear $\mathrm{LBO}$ crystal $\left({\mathrm{LiB}}_3{\mathrm{O}}_5\right)$. The UV pulses pass through a 2-mm $\mathrm{BBO}$ ($\beta$-barium borate ) crystal and polarization entangled photon pairs in the state, $\mid {\psi }^{-}{⟩}_{12}$ are created. In order to compensate the birefringence of the $\mathrm{BBO}$ crystal, we place a half-wave plate (HWP) and a compensating $\mathrm{BBO}$ crystal of $1\mathrm{mm}$ thickness on each path of the two photons as a compensating system (labeled CS). Three electro-optic modulators controlled by a random number generator are utilized to produce the other three two-photon states requested by the protocol. They are labeled M1–M3, respectively. The channel noise of random rotation is realized by four half-wave plates, two in each path ($H_C$ and $H_N$). We use an electro-optic modulator M4 controlled by another random number generator to choose the measurement bases. After the modulator, we place polarized beam splitters (PBS), and the raw key is obtained by observing the clicking of detectors behind the PBS. An interference filter (IF) with a full width at half maximum of $2.8\mathrm{nm}$ is placed before each single photon detector (D1–D4) to improve the visibility of the entangled pair. Each user has a computer to control the random number generator and record the detector’s events.Reuse & Permissions

Figure 2

(Color online) (a) Experimental result of the quantum bit error rates of all the states in our protocol and the BB84 protocol in the rotation noise. (b) Experimental result of the total quantum bit error rate of our protocol and the BB84 protocol. It can be seen at any angle that our protocol is below the line of 11%, which is the security bound of the BB84-type protocol. The error rates of our protocol are due to the imperfection of the entanglement resource, the detectors, and the electro-optic modulators. The error rates of the BB84 protocol are sinusoidal and are the same to our protocol at 0° angle rotation noise, which are in agreement with theoretical prediction.Reuse & Permissions

Figure 3

(Color online) Coincidence fringes for the entangled photon source in our experiment whose state is $\mid {\psi }^{-}{⟩}_{12}$. When varying the polarizer angle ${\theta }_1$, the two complementary sine curves with a visibility of 88%, will bring a 6% error rate to the QKD protocol, which is the main reason of the QBER of our experiment.Reuse & Permissions