Title: "The Design Philosophy of the DARPA Internet Protocols".
Author: Yu Liao
Problem:
The motivations, reasonings, phisolophies behind the design of the DARPA Internet Protocols
Contribution
Through illustrating the Fundamental goal, and cataloging the second-level goals, the authors reason about the design of the Internet Protocols, how tradeoff is made to achieve the goals, these reasoning is helpful in understanding the Internet Protocols, why they're as they are today, on the other hand, it also helps guide designers design new network protocols, or make improvements over existing protocols.
Key Ideas
- Based on the fundamental goal: Effectively multiplex utilization of existing interconnected networks, the fundamental structure of the Internet comes as: A packet switched communications facility in which a number of distiguishable networks are connected using gateways which implement a store and forward packet forwarding algorithm.
- The design goals are ordered based on its relative importance, survivability, support multiple types of services and support a variety of networks are considered of paramount importance, and they decide the provision of datagram as a basic building block, and the layering of TCP and IP.
- The Internet architecture provide wide flexibility in the services offered, at the same time, it leaves the designer of a particular realization with a great deal of engineering to do.
Critique
- modest contribution: since it's just a summarization of the reasonings made in the design of the Internet protocols.
- a new basic building block called "flow" is proposed in the paper, and the concept "soft state" for dealing with the lower end of the goals.
Lessons learned
Reasearchers should prioritize the goals, and tradeoff the great variety of designs to meet the goals. It's a mistake to only attend to the functionality, while ignoring the performance issue, but how to formalize any apspect of performance constraints within the architecture is chllenging, and left open in the paper.