Today’s Internet architecture is facing significant challenges posed by an ever increasing array of devices and applications, tremendous amount of traffic, as well as serious security threats. There has been considerable debate in the networking community on how to evolve or redesign the Internet.
However, our graduate networking courses often favor the mechanical analysis of specific protocol details over understanding of high-level architectural ideas. As a result, our students master the how, often at the expense of the why.
This paper presents our attempt to add more examination of architectural designs into graduate courses on networking. Our premise is that a network architecture is not borne out of a singular piece of work, but rather a progression of ideas that can be traced through a series of papers. We believe it is important for our students to understand the progression of these ideas and the design tradeoffs being made, rather than simply knowing how the architecture works.
To illustrate our approach while limiting new material, we focus on two network architectures, the current Internet and one of the new architectures based on content-centric networking. For each architecture we provide a series of five papers that aim to show both the big ideas and the evolution of their architectural designs.