Fig. 2: Numerical calculation of WI distribution and magneto-conductance in a sample with antidot defects.

a A sample system for calculation with two antidot defects. One antidot is fixed at the upper centre of the nanowire, while the other is located so as not to overlap with the fixed one. Two leads are attached to the top and bottom ends of the sample to measure conductance. b Calculated WI distribution. The square of the absolute value of the calculated wave function is plotted in the sample region with 60 \({\times}\) 50 pixels. The WI values are normalised such that the sum of the intensity equals to unity for each antidot configuration; \({\sum }_{i=1}^{60}{\sum }_{j=1}^{60}{X}_{i,j}=1\), where \({{{\bf{X}}}}\) is a WI image, and the suffixes \(i,{j}\) represent the pixel label. We added zero padding with 5 sites to the left and right ends of the nanowire. c Magneto-conductance \(G\) for 10 samples with different defect distributions. Here, B is the magnetic flux density. d Normalised data of the calculated magneto-conductance \(\Delta G\) for the antidots distribution shown in a. \(\Delta G\) is obtained by subtracting the averaged \(G\ (\equiv {G}_{{{{{{\rm{ave}}}}}}})\) over all the nanowire configurations from the \(B\) dependence of \(G\ [\Delta G\left(B\right)\equiv G\left(B\right)-{G}_{{{{{{\rm{ave}}}}}}}(B)]\).