A combined TDA/FDA adaptive schema for stereophonic acoustic echo cancellation

In many applications, such as teleconferencing, multimedia workstations, tele-video gaming, etc., stereo sound is implemented to give spatial realism that mono-channels cannot offer. In hands-free stereophonic teleconferencing systems, the loudspeaker signals are highly correlated, leading to slow convergence rates for the gradient algorithms used. In this paper, we propose two different types of adaptation; one for each channel. Simulation results are presented to show that hybrid-type adaptations can be used in practical implementations to overcome the impact of large eigenvalue spreads, due to channel correlations, which result in increased speeds of convergence for the underlying adaptive algorithms.