A study of multichannel adaptive lattice algorithms as applied to the equalization of nonlinear channels is presented. A single-input-multiple-output system is proposed to be used in the equalization model. This system extends the nonlinear channel output as a vector. Since the single-input-multiple-output system gives an output which has a second-order nonlinearity, the nonlinear channel can be taken as in the form of a second-order Volterra filter. The performance of the algorithm is examined and it is shown that in a few iterations, the equalizer estimates the channel input correctly.