This paper presents a multiple sub-filter-based improved nonlinear acoustic echo cancellation (NAEC) framework to enhance the echo cancellation performance of the NAEC in the presence of nonlinear distortion. The proposed algorithm uses a novel combination of adaptive multiple sub-filter approach and proportionate filtering to enhance the convergence rate of the existing proportionate functional link-based NAEC algorithm by reducing the adaption time needed for updating the coefficients of the adaptive filter. In addition to that, the convergence and the steady-state analysis of the proposed algorithm are presented. The proposed NAEC framework is subjected to speech signal input corrupted with both white as well as colored noise as background noise and the colored noise input at low-to–high SNR conditions for a comprehensive analysis of the echo cancellation performance. The experimental results comprising of the echo return loss enhancement, spectrograms, and the perceptual evaluation of speech quality demonstrate the improvements brought by the proposed algorithm. At all signal-to-noise ratio conditions, the proposed algorithm has shown a 4 dB improvement in mean echo return loss enhancement compared to the existing algorithms validating the improvement in the proposed NAEC scheme's performance.