Abstract-In this paper, we present the expressions, not published up to now, that describe the AM/AM and AM/PM conversions of communication power amplifiers (PAs) via the Volterra series based nonlinear transfer functions. Furthermore, we present a necessary and sufficient condition of occurrence of the nonzero values of AM/PM conversion in PAs. Moreover, it has been shown that Saleh's approach and related ones, which foresee nonzero level of AM/PM conversion, are not models without memory. It has been also shown that using a polynomial description of a PA does not lead to a nonzero AM/PM conversion. Moreover, a necessary condition of occurrence of an AM/AM conversion in this kind of modelling is existence of at least one nonzero polynomial coefficient associated with its odd terms of degree greater than one.
This paper presents how to find an architecture for very large scale lossless neural nets, which can he used as Haar-Walsh spectrum analyzers. This analysis relies on the orthogonality of weight matrices W, where W could be Hurwitz-Radon matrices. The unique feature of these nets is the possibility to treat them either as algorithms or as Hamiltonian physical objects (Haar-Walsh Signal Processors).
We point out that Hamiltonian Neural Networks (HNN) and based on HNN orthogonal filters can be used as universal signal processors. To illustrate it, we propose a procedure for design of large scale mappings, classifiers and supervised learning systems.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.