We present a way to estimate the aerosol complex refractive indices (ACRI) based on a real-time Stokes vector measurement technology. First, we introduce how to simultaneously get the multi-polarization signals of multi-scattering angles. Then we demonstrate the numerical inversion to retrieve an ACRI based on an iterative Mie algorithm. Meanwhile, we show the experimental results of several types of aerosol samples. Our optimal estimation of ACRI inversion shows a good agreement with the references, which confirms the feasibility and then implies a good prospect of multidimensional polarization characterization applied in the future aerosol recognition, especially suitable for near-spherical scatters.