Fig. 1. (a) Experimental setup of generating photon pairs from an N-stage nonlinear interferometer (NLI) consisting of N pieces of nonlinear media of dispersion shifted fibers (DSFs) and N−1 pieces of phase shifters of standard single mode fibers (SMFs). DBF, dual-band filter; FPC, fiber polarization controller; FPBS, fiber polarization beam splitter; SPD, single photon detector; DAQ, data acquisition system. (b) The contour of JSI |F(ωs,ωi)|2 for photon pairs generated from a single piece of DSF.

Fig. 2. Results for N-stage NLI with N = 2, 3, 4 when the lengths of each DSF and SMF in the NLI are 100 m and 10 m, respectively. (a), (b) The contours of interference factor |H(θ)|2 and JSI |FNLIS(ωs,ωi)|2 calculated by substituting experimental parameters into Eqs. (2) and (1), respectively. (c) Contour maps of measured true coincidence of photon pairs, which reflect the JSI of photon pairs. (d) Measured (dots) and calculated (solid curves) marginal intensity distributions of signal photons.

Fig. 3. Measured counting rate (dots) R_s as a function of the average pump power P_a for the photons in an individual signal band centering at (a) 1560.4 nm and (b) 1561.9 nm, respectively. The solid curve Rs=s1Pa+s2Pa2 is the fitting function of raw data (dots); dashed and dotted lines are the linear term s1Pa and quadratic term s2Pa2, respectively.

Fig. 4. Results for uneven N-stage NLI, in which the lengths of DSFs, labelled in the top for different stage number N, follow binomial distribution. (a), (b) The calculated contours of interference factor |K(θ)|2/L12 and JSI |FNLI(UN)(ωs,ωi)|2 [see Eqs. (5) and (6)], respectively. (c) Contour maps of true coincidence of photon pairs, reflecting the JSI of photon pairs. (d) Measured (dots) and calculated (solid curves) marginal intensity distributions of signal photons.