Multi-level modeling in SARA

  • Authors:
  • Robert I. Gardner

  • Affiliations:
  • Computer Science Department, University of California, Los Angeles, California

  • Venue:
  • Proceedings of the Symposium on Design Automation and Microprocessors
  • Year:
  • 1977

Quantified Score

Hi-index 0.00

Visualization

Abstract

Each model of an evolving system in SARA is a representation of an abstraction of the structure or behavior of that system. The degree or level of abstraction of a model is chosen such that enough detail is represented so that the model provides adequate accuracy for the purposes for which it will be used but not so much detail that the model becomes unduly complicated, difficult to build or difficult to simulate. The technique of multi-level modeling allows a system to be modeled at more than one level of abstraction to provide both detailed modeling and a limit to the complexity of the model [ZUR68]. In top-down design, each successively lower-level model encompasses more detail and possibly less scope than the previous models and in bottom-up design each successively higher-level model encompasses less detail and possible more scope than the previous models. Two different approaches to multi-level modeling that have been used in the implementation of SARA are presented here. One approach is used in the SLI structural modeling tool and allows system structures to be represented as multi-level hierarchies of nested modules. The second approach is used in the GMB behavioral modeling tool and allows the designer to easily develop higher- and lower-level GMB models but represents each model as a single-level model.