Matrix Analysis of Mildly Nonlinear, Multiple-Input, Multiple-Output SystemsWith Memory

Communication, control, and instrumentation systems employ components, such as amplifiers and mixers, which are inherently nonlinear. Even when the nonlinearities are mild, as is often the case, they can produce bothersome signal distortion that limits the system performance. The nonlinear components themselves, and the other linear components used in the system, are generally frequency-dependent, i.e., they have memory. Numerous studies are available in the literature for the analysis of mildly nonlinear systems with memory through the use of Volterra-series expansions.1-21 The classic paper by Bedrosian and Rice,7 the recent paper by Chua and Ng,14 and the book by Weiner and Spina 1 ' cover that subject very thoroughly. Also, the paper by Gopal, Njakhla, Singhal, and Vlach12 is interesting in that it evaluates the range of accuracy of the Volterra-series approach by comparing it with a nearly exact, but quite involved, method of analysis. The book18 and paper19 by Schetzen deal mainly with random inputs. The condi2221