MULTICHANNEL ADAPTIVE LEAST SQUARES LATTICE FILTERS FOR ACTIVE NOISE CONTROL

In this paper a new algorithm for multi-channel adaptive least squares adaptive filtering, using a lattice filter is pro- posed. The algorithm is a multi-channel numerically sta- ble version of the adaptive least squares lattice with a pri- ori errors and error feedback. It is adapted to active noise control using the modified filtered-Error algorithm. It is shown throw computer simulations that the algorithm is sta- ble, and that it has better convergence properties than algo- rithms based on the LMS and gradiente adaptive lattice.