High level modeling of signal integrity in field bus communication with SystemC-AMS

Abstract

This paper presents a novel method for modeling the functionality of a mixed-signal system, and analyzing its signal integrity (SI) at a high-level of abstraction with SystemC-AMS. Our model includes on a unique platform a functional module and a non-functional module. The functional module represents the operative behavior of the system and the nonfunctional module, based on neural network techniques, displays the SI characteristics of the system. The proposed method is demonstrated by modeling field bus communication system with two nodes. We achieved an error of about 3% for the neural network based Time Data Flow (TDF) model with respect to a RLC Electrical Linear Networks (ELN) model. IndexTerms -High-level modeling, signal integrity, field bus, SystemC-AMS.

High-Level Virtual Prototyping of Signal Integrity in Bus Communication

“…SystemC-AMS offers two versions of TDF: conventional TDF [21], in which data are sampled with a fixed time step, and dynamic TDF [31], in which the time step can be dynamically changed. Compared with a conventional TDF module, a dynamic TDF module samples and computes less data, thus achieving a potential simulation speedup [32]. This feature will be illustrated in Section III-D.…”

“…The first two stages were already introduced in [32]. Input/target pairs were obtained by the simulation of an equivalent circuit with NGSPICE (Fig.…”