Paper Review: < BCS94 >
Integrated Services in the Internet Architecture: an Overview

Reviewer: Oleg Elkhunovich

Problem

This paper purposes an extension framework for the Internet to be better suited to provide a wide variety of applications, dealing with real-time traffic.

Contribution

Paper explored the shortcomings of the existing Internet architecure in response to demand for real-time applications. It purposes an extension framework to support those applications better: Integrated Services.

Main Ideas

  • It is impossible to manage resources efficiently at the IP level. Existing protocols are not suited well for real-time applications. IS model is meant to provide different kinds of service for different applications depending on the needs. Real-time QoS is necessary in this case.
  • IS framework consists of packet scheduler, the admission control routine, the classifier, and the reservation setup protocol. The coordination takes place at the routers.
  • Paper purposes a single service model for the Internet

    Critique

    Significance: 3
    Paper addresses a very important topic and purposes an interesting solution to the existing problem. From the scientific point of view however, paper does not present any evidence to back-up its claims (expected as this is RFC).
    Methodology:
    Methodology is not very good from the point of view of the research paper as it does not present many experiements or scientific analysis. However, it is important to realize that this is a Request for Comment document and not a publishable research paper. It is interesting to compare and contrast these two different types of work.
    Limitation:
    The framework purposed seems to be very complex and possible not feasible to implement. A lot of major changes in the architecure would be required. Seems like authors want revolution more than evolution.

    Lessons:
    Real time applications are very important for the future but solutions have not yet emerged to this complex problem.