On the way to a distributed systems calculus: an end-to-end network calculus with data scaling

  • Authors:
  • Markus Fidler;Jens B. Schmitt

  • Affiliations:
  • Centre for Quantifiable Quality of Service, NTNU Trondheim, Norway;University of Kaiserslautern, Germany

  • Venue:
  • SIGMETRICS '06/Performance '06 Proceedings of the joint international conference on Measurement and modeling of computer systems
  • Year:
  • 2006

Quantified Score

Hi-index 0.00

Visualization

Abstract

Network calculus is a min-plus system theory which facilitates the efficient derivation of performance bounds for networks of queues. It has successfully been applied to provide end-to-end quality of service guarantees for integrated and differentiated services networks. Yet, a true end-to-end analysis including the various components of end systems as well as taking into account mid-boxes like firewalls, proxies, or media gateways has not been accomplished so far. The particular challenge posed by such systems are transformation processes, like data processing, compression, encoding, and decoding, which may alter data arrivals drastically. The heterogeneity, which is reflected in the granularity of operation, for example multimedia applications process video frames which, however, are represented by packets in the network, complicates the analysis further. To this end this paper evolves a concise network calculus with scaling functions, which allow modelling a wide variety of transformation processes. Combined with the concept of packetizer this theory enables a true end-to-end analysis of distributed systems.