Router Scheduling Configuration Based on the Maximization of Benefit and Carried Best Effort Traffic

Short Description
This paper shows a configuration scheme for networks with WFQ schedulers. It guarantees maximum revenue for the service provider in the worst case of network congestion. We focus on best effort traffic and select those flows that maximize the benefit while keeping the network utilization high. We show that optimum network configuration is feasible based only on knowledge of the topology. Its dependence on the pricing scheme can be reduced and even eliminated. We offer a formulation that reaches a tradeoff between network utilization, fairness, and user satisfaction.

Website: paleale.eecs.berkeley.edu | Filesize: 159kb
No of Page(s): 18

Content
Providing Quality of Service (QoS) requirements for certain flows in “best effort” IP networks is a topic of attention from researchers, enterprises and Internet Service Providers (ISPs).
Solutions based on DiffServ (Differentiated Services) [Liebeherr and Christin, 7] or IntServ (Integrated Services) [White, 18] provide mechanisms to guarantee certain throughput and delay to the flows with QoS constraints in an individual autonomous system [Xiao and Ni, 19]. They focus on two classes of traffic: flows with quality of service requirements (that we will call EF or Expedited Forwarding) and Best Effort (BE) traffic.
For the provision of this QoS, new schedulers have been implemented in network routers. Schedulers like Weighted Fair Queuing (WFQ) [Demers et al., 3], Packetized Generalized Processor Sharing (PGPS) [Parekh and Gallager, 13] and Class Based Queuing (CBQ) [Floyd and Jacobson, 5] can provide a minimum bandwidth for required flows. The configuration of the schedulers is straightforward from the requirements of the EF flows if we use their bandwidths as the weights in the scheduler [Parekh and Gallager, 13]. However, these routers typically use this scheduling mechanism with BE traffic too. The default configuration gives the same weight to every flow or a weight based on the TOS bits in the IP header. There is a lack of an accepted solution for the configuration of weights for these flows without requirements, a solution that could be applied to the huge variety of services and traffic types found in data networks. Even for the flows from services that carry a large percentage of the network traffic, it is not easy to optimize their impact on the network.
…

Get the file Download here