The stream machine: a data flow architecture for real-time applications

  • Authors:
  • Paul Barth;Scott Guthery;David Barstow

  • Affiliations:
  • Schlumberger-Doll Research, Old Quarry Road, Ridgefield, Conn;Schlumberger Well Services, 12125 Technology Blvd., Austin, Texas;Schlumberger-Doll Research, Old Quarry Road, Ridgefield, Conn

  • Venue:
  • ICSE '85 Proceedings of the 8th international conference on Software engineering
  • Year:
  • 1985

Quantified Score

Hi-index 0.00

Visualization

Abstract

The Stream Machine is a software architecture designed to support the development and evolution, as well as the efficient execution, of software that performs both data acquisition and process control under real-time constraints. A stream machine program consists of a set of concurrently executing modules communicating through streams of data. Streams provide essentially a data flow style of communication, thereby supporting deterministic data acquisition and calculation. The basic constructs have been augmented by several time-based operations to support process control software. In addition, explicit declarations of timing constraints layered on Stream Machine programs are currently being explored for resource allocation and scheduling. Our experience to date with two implementations of the Stream Machine suggests that it facilitates the mixture of software that performs both data acquisition and process control.