A Representative Application of a Layered Interface Modeling Pattern

Abstract

AbstractModel‐based systems engineering (MBSE) is intended to improve how systems engineering is performed compared with a more traditional document‐based approach by effectively using models to analyze, specify, design, and verify systems. The OMG Systems Modeling Language (OMG SysML™) enables the practice of MBSE by providing a robust and expressive language for representing systems.Several MBSE methods are available [3], and have continued to mature over the last several years which include model‐based practices for requirements flow‐down, architecture‐design, trade‐off analysis, verification planning, and others. One of the critical systems engineering practices is interface modeling. This paper describes a layered interface pattern for modeling data and communications interfaces using SysML. The pattern spans logical to physical interface definition, and includes software and electrical interfaces.Each layer in a stack describes a portion of the interface functionality. The concept of a layered interface is borrowed from computer networking [8] [10]. The layered interface pattern described in this paper enables the specification and design of connections and behavior between interfacing systems at a given layer, and between the adjacent layers of a single system. This pattern may also be applied recursively. That is, communication within a single layer may itself be realized by a multi‐layer stack. The level of detail of the model to describe a layered interface should be adapted to the need, and can vary from highly abstract logical flows across a system to highly detailed protocol specifications and message structures.This paper builds on work that was documented in a previous paper entitled “A modeling pattern for layered system interfaces” [5]. Aspects of this pattern have been demonstrated in various project applications including Exploration Flight Test 1 (EFT‐1), Space Communication and Navigation (SCaN) Trade Studies, and the SCaN Network Integration Project (SNIP).