Proportional differentiated services

書誌事項

タイトル別名
  • delay differentiation and packet scheduling

抄録

<jats:p> Internet applications and users have very diverse service expectations, making the current <jats:italic>same-service-to-all</jats:italic> model inadequate and limiting. In the <jats:italic>relative differentiated services</jats:italic> approach, the network traffic is grouped in a small number of <jats:italic>service classes which are ordered based on their packet forwarding quality</jats:italic> , in terms of per-hop metrics for the queueing delays and packet losses. The users and applications, in this context, can <jats:italic>adaptively</jats:italic> choose the class that best meets their quality and pricing constraints, based on the assurance that <jats:italic>higher classes will be better, or at least no worse, than lower classes</jats:italic> . In this work, we propose the <jats:italic>proportional differentiation model</jats:italic> as a way to refine and quantify this basic premise of relative differentiated services. The proportional differentiation model aims to provide the network operator with the ' <jats:italic>tuning knobs</jats:italic> ' for adjusting the quality spacing between classes, <jats:italic>independent of the class loads</jats:italic> ; this cannot be achieved with other relative differentiation models, such as strict prioritization or capacity differentiation. We apply the proportional model on queueing-delay differentiation only, leaving the problem of coupled delay and loss differentiation for future work. We discuss the <jats:italic>dynamics</jats:italic> of the proportional delay differentiation model and state the conditions under which it is <jats:italic>feasible</jats:italic> . Then, we identify and evaluate (using simulations) two <jats:italic>packet schedulers</jats:italic> that approximate the proportional differentiation model in heavy-load conditions, even in short timescales. Finally, we demonstrate that such per-hop and class-based mechanisms can provide consistent end-to-end differentiation to <jats:italic>individual flows</jats:italic> from different classes, independently of the network path and flow characteristics. </jats:p>

収録刊行物

被引用文献 (11)*注記

もっと見る

詳細情報 詳細情報について

問題の指摘

ページトップへ