Towards an evolvable internet architecture

Abstract

There is widespread agreement on the need for architectural change in the Internet, but very few believe that current ISPs will ever effect such changes. In this paper we ask what makes an architecture evolvable, by which we mean capable of gradual change led by the incumbent providers. This involves both technical and economic issues, since ISPs have to be able, and incented, to offer new architectures. Our study suggests that, with very minor modifications, the current Internet architecture could be evolvable.

Innovations and upgrades in virtualized network architectures

“…Able to choose any service provider it wants, it should come as no surprise that a network provider will often plump for itself, and avoid having to split profits with another firm. For example, Ratnasamy et al consider the possibility of network evolution through competing network layer protocols [13]. The IP layer network can certainly be considered as a service in our model, and some users might prefer, say, IPv6 over IPv4.…”

“…This is the "hard case" for deployment, because a technology does not have much value if only a portion of providers adopt it. This implies that a small group of adopters cannot gain much competitive advantage over their rivals (such competitive advantages are used to support adoption in [4][10] [13]). On the other hand, the technology will achieve its maximum value if all providers adopt it, but then no firm has any competitive advantage so profits may not rise by much.…”

“…Unfortunately, even as our aspirations have risen, the network in use by the vast majority of users has remained stagnant, and even ossified [3] [12]. Many studies have noted the internet's resistance to evolution [3][5] [8][10] [11][12] [13][16] [17]. Internet service providers are the gatekeepers to any architectural change, and we now accept that they lack the economic incentives to adopt new technologies [10] [13].…”

A survey on the design, applications, and enhancements of application-layer overlay networks

“…In this article, we aim to provide the reader with a comprehensive overview of the more significant applications for overlay networks, the issues that arise with each of these applications and some existing solutions to these issues. We also present recently proposed two main views, called the purist [Ratnasamy et al 2005] and pluralist argument ], on the long-term impact of overlay networks in the Internet. While the purist view sees overlay networks as testbeds for experimentation with novel network architectures, the pluralist view considers overlays as an integral part of the future Internet in providing value-added network services to end users.…”

OpenFlow and Xen-Based Virtual Network Migration

“…This problem causes a dead lock, on the one hand Internet must evolve to handle new demands, but on the other hand, new mechanisms that change the Internet cannot be applied in the underlying infrastructure. Many works tackle this problem by using virtualization techniques [4,9] because, in a virtualized environment, the physical substrate is shared among many virtual networks. Virtual networks are isolated and, consequently, new proposals could be run together with the current Internet without disturbing the production traffic.…”