Convergence behavior of two adaptive filters in tandem

The current communication networks often encounter echo-cancellers in tandem. The major component of an echo-canceller is an adaptive filter, which learns the echo-path impulse response dynamically to generate an echo estimate for cancellation. Tandeming affects the learning behavior of adaptive filters in the cancellers and may create a stability problem. At present there is little work in the literature to investigate the behavior of adaptive filters in tandem. This paper presents a theoretical study of the learning characteristics when two adaptive filters are connected together. The adaptive algorithm is assumed to be LMS. The theoretical results are corroborated with simulations.