Today's Internet provides one simple service: best effort datagram delivery. This minimalist service allows the Internet to be stateless, that is, routers do not need to maintain any fine grained information about traffic. As a result of this stateless architecture, the Internet is both highly scalable and robust. However, as the Internet evolves into a global commercial infrastructure that is expected to support a plethora of new applications such as IP telephony, interactive TV, and e-commerce, the existing best effort service will no longer be sufficient. In consequence, there is an urgent need to provide more powerful services such as guaranteed services, differentiated services, and flow protection.

Over the past decade, there has been intense research toward achieving this goal. Two classes of solutions have been proposed: those maintaining the stateless property of the original Internet (e.g., Differentiated Services), and those requiring a new stateful architecture (e.g., Integrated Services). While stateful solutions can provide more powerful and flexible services such as per flow bandwidth and delay guarantees, they are less scalable than stateless solutions. In particular, stateful solutions require each router to maintain and manage per flow state on the control path, and to perform per flow classification, scheduling, and buffer management on the data path. Since today's routers can handle millions of active flows, it is difficult, if not impossible, to implement such solutions in a scalable fashion. On the other hand, while stateless solutions are much more scalable, they offer weaker services.

The key contribution of this dissertation is to bridge this long-standing gap between stateless and stateful solutions in packet switched networks such as the Internet. Our thesis is that ``it is actually possible to provide services as powerful and as flexible as the ones implemented by a stateful network using a stateless network''. To prove this thesis, we propose a novel technique called Dynamic Packet State (DPS). The key idea behind DPS is that, instead of having routers maintain per flow state, packets carry the state. In this way, routers are still able to process packets on a per flow basis, despite the fact that they do not maintain any per flow state. Based on DPS, we develop a network architecture called Stateless Core (SCORE) in which core routers do not maintain any per flow state. Yet, using DPS to coordinate actions of edge and core routers along the path traversed by a flow allows us to design distributed algorithms that emulate the behavior of a broad class of stateful networks in SCORE networks.

In this dissertation we describe complete solutions including architectures, algorithms and implementations which address three of the most important problems in today's Internet: providing guaranteed services, differentiated services, and flow protection. Compared to existing solutions, our solutions eliminate the most complex operations on both the data and control paths in the network core, i.e., packet classification on the data path, and maintaining per flow state consistency on the control path. In addition, the complexities of buffer management and packet scheduling are greatly reduced. For example, in our flow protection solution these operations take constant time, while in previous solutions these operations may take time logarithmic in the number of flows traversing the router.

209 pages

Return to: SCS Technical Report Collection
School of Computer Science homepage

This page maintained by [email protected]