In response to the boundary value issue of the extinction coefficient in single-channel aerosol lidar inversion technology, an aerosol extinction coefficient inversion algorithm based on exponential iteration has been proposed to optimize the boundary value in the inversion calculation. The 4th-order convergent exponential iteration method is utilized to solve the nonlinear equation between the extinction coefficient boundary value and the backscattering signal of the lidar, and the Fernald method is used to invert the aerosol extinction coefficient. By using exponential iteration combined with the Fernald method, horizontal scanning data and vertical fixed-point detection data were inverted. The results were compared with those obtained from the Klett and Mie-Raman methods, which served as reference values to validate the performance of the exponential iteration method. The data analysis results indicate that, compared to the fixed-point iteration method, the exponential iteration method exhibits higher precision, faster convergence, and fewer iterations. The inversion error is approximately 14%, with around five iterations taking about 9 μs. This approach enables the rapid and accurate inversion of the extinction coefficient boundary values for horizontal scanning and vertical fixed-point detection signals from lidar.