Abstract

Quality of Service indicators in computer networks reached tremendous importance over the last years. Especially throughput and latency are directly influenced by the dimension of packet queues. Determining the optimal dimension based on the inevitable tradeoff between throughput and latency tends to be a hard, almost infeasible challenge. Several algorithms for Active Queue Management have been proposed to address this challenge over the last years. However, the deployment and by that the development of such algorithms is challenging as they are usually located within the operation systems’ kernel or implemented in fixed hardware. In this work, we investigate how novel algorithms can be deployed in user space for rapid prototyping with tolerable effort. We provide core performance characteristics and highlight the viability and reasonability of this approach.

Full Text
Paper version not known

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call