Title :
Real-time echo cancellation using a new fast LMS-based algorithm
Author :
Khasawneh, M.A. ; Haddad, T.F.
Author_Institution :
Dept. of Electr. Eng., Jordan Univ. of Sci. & Technol., Irbid, Jordan
Abstract :
We present a real-time implementation of a new fast LMS-based algorithm, which was shown in earlier work to supersede the LMS algorithm in its rate of convergence with little misadjustment. Our implementation employs the new algorithm in echo cancellation in a telephone network with the TMS32020 processor. The implementation shows the advantage of the new algorithm over the conventional LMS in achieving faster convergence with less misadjustment, due to the larger bound allowed on the step size, μ, derived for the algorithm. This result is demonstrated in simulating the operation of the echo canceller using actual noisy speech signals
Keywords :
digital signal processing chips; echo suppression; least mean squares methods; telecommunication computing; telephone networks; TMS32020 processor; convergence rate; fast LMS-based algorithm; noisy speech signals; real-time echo cancellation; step size; telephone network; Convergence; Echo cancellers; Finite impulse response filter; Least squares approximation; Noise cancellation; Signal processing; Signal processing algorithms; Speech; Telephony; Working environment noise;
Conference_Titel :
Acoustics, Speech, and Signal Processing, 1994. ICASSP-94., 1994 IEEE International Conference on
Conference_Location :
Adelaide, SA
Print_ISBN :
0-7803-1775-0
DOI :
10.1109/ICASSP.1994.390050