Autonomous end to end QoS monitoring

Abstract

Verifying that each flow in the network satisfies its QoS requirements is one of the biggest scalability challenges in the current DiffServ architecture. This task is usually performed by a centralized allocation entity that monitors the flows' QoS parameters. Efficient detection of problematic flows is even more challenging when considering aggregated information such as the end to end delay suffered by packets belonging to a specific flow. Known oblivious and reactive monitoring techniques do not scale well when the number of flows and the length of their paths increase, and when the network load increases. This is due both to load on the centralized bandwidth allocation entity and to the excessive number of monitoring and control messages needed. We propose a new monitoring paradigm termed autonomous monitoring, in which the network itself (i.e. the routers along the flow path) is responsible to discover when a violation of the SLA occurs (or is soon to occur). Only in such cases the centralized allocation entity is notified, and can take the required actions. We study the performance of this new distributed algorithm through theoretical analysis and extensive simulations. Our results indicate that in addition to dramatically reducing the load from the centralized allocation entity, the amount of network traffic needed is relatively small and thus the new monitoring scheme scales well.