Fig. 1. Structure schematic of a 4-bit binary PSA

Fig. 2. Flowchart of the iteration and optimization process

Fig. 3. Simulation on the sensitivity characteristics of the proposed optimization method at different states of polarization (SOPs). (a) The variation slope of I16×122 related to φ; (b) the variation slope of I16×122 related to θ_p; (c) the variation slope of I16×122 related to Γ

Fig. 4. Simulation flow of the accuracy evaluation of self-calibration and SOP analysis based on Monte Carlo analysis

Fig. 5. Self-calibrated AAE distributions at different SOPs. (a) AAE of φ; (b) AAE of θ_p; (c) AAE of Γ

Fig. 6. Comparison in accuracy of the two methods at four SOP points. (a)‒(c) CDF of S₁, S₂, and S₃ at SOP1; (d)‒(f) CDF of S₁, S₂, and S₃ at SOP2; (g)‒(i) CDF of S₁, S₂, and S₃ in SOP3; (j)‒(l) CDF of S₁, S₂, and S₃ in SOP4

Fig. 7. Comparison in AAE of the two methods. (a) AAE of S₁; (b) AAE of S₂; (c) AAE of S₃; (d) AAE of DOP

Fig. 8. Comparison in STD of the two methods. (a) STD of S₁; (b) STD of S₂; (c) STD of S₃; (d) STD of DOP

Fig. 9. Schematic and photographs of the experimental systems. (a) Schematic of the experimental system; (b) the photograph of the experimental system with PLM measuring SOPs; (c) photograph of the experimental system with PM measuring SOPs

Fig. 10. Experimental results on self-calibration, SOP analysis, and DOP. (a) Self-calibration of φ; (b) self-calibration of θ_p; (c) self-calibration of Γ; (d)‒(f) SOP analysis of S₁, S₂, and S₃ obtained by the proposed and direct methods in four SOPs; (g) DOP obtained by the proposed and direct methods in four SOPs

Fig. 11. Experiments on the wavelength dependence. (a) Wavelength dependence of φ; (b) wavelength dependence of θ_p; (c) wavelength dependence of Γ