A low complexity Hirschman Optimal Transform based feedback cancellation scheme for hearing aid

The hearing aid being a portable, compact, battery operated device, requires algorithms with low computational complexity. The adaptive feedback canceller (AFC) based on the time domain least mean square (LMS) algorithm exhibits slow convergence and degraded performance when colored signal is used as input and also suffers from large computational complexity. The frequency domain block LMS (FBLMS) algorithm performs better in terms of convergence, steady state performance and computational complexity than the time domain LMS algorithm for colored signal as input. However, the FBLMS algorithm suffers from heavy computational complexity due to the large size of the DFT involved while modelling the feedback path of the hearing aid. To overcome this problem, a new structure using the Hirschman Optimal Transform (HOT)-DFT LMS (HDFT) algorithm is developed as adaptive feedback canceller (HDFT-AFC) for hearing aids. The simulation results demonstrate that the proposed method offers significant computational savings than the FBLMS algorithm without any performance degradation when used for feedback cancellation in hearing aids.