The Forwarding on Gates Architecture: Merging IntServ and DiffServ

Abstract—Quality of Service (QoS) will be a major enabler for Future Internet applications and services. However, today’s Internet provides no suitable QoS support for end-to-end connections due to several drawbacks of IntServ and DiffServ. Therefore, this paper proposes the “Forwarding on Gates” architecture, which uses a new network protocol designed to handle IntServ and DiffServ in an integrated way. The architecture supports resource reservations for QoS guarantees, like in IntServ scenarios, and prioritized traffic, like in DiffServ scenarios. Furthermore, a combination of both is supported. This paper introduces the architecture and defines the network protocol used to implement these features. The evaluation includes theoretical descriptions of the network configuration for the different scenarios and simulation results concerning the protocol overhead in large-scale networks. Our new architecture is able to support QoS in a scalable way, since it allows a network providing QoS to move states and delegate decisions about the QoS usage to the entity using the QoS. Keywords—Future Internet; network protocol; architecture; QoS