Adaptive algorithms for identifying recursive nonlinear systems

ABSTRACT This paper presents two fast least-squares lattice algorithms for adaptive non-linear filters equipped with system models involving nonlinear feedback. Such models can approximate a large class of non-linear systems adequately, and usually with considerable parsimony in the number of coefficients required. For simplicity of presentation, we consider the bilinear system model in the paper, even though the results are applicable to more general system models. The computational complexity of the algorithms is an order of magnitude smaller than previously available methods. Results of several experiments that demonstrate the properties of the adaptive bilinear filters as well as compare their performances with two other algorithms that are computationally more expensive are also presented in this paper.