DIBS: A dual interval bandwidth scheduling algorithm for short-term differentiation

Abstract

Packet delay and bandwidth are two important metrics for measuring quality of service (QoS) of Internet services. While proportional delay differentiation (PDD) has been studied intensively in the context of differentiated services, few studies were conducted for per-class short-term bandwidth differentiation. In this paper, we design and evaluate an efficient bandwidth differentiation algorithm. The dual interval bandwidth scheduling (DIBS) algorithm focuses on the short-term bandwidth differentiation of multiple classes because many Internet transactions take place in a small time frame. It does so based on the normalised instantaneous bandwidth, measured by the use of packet size and packet delay. It also uses two intervals to trade off differentiation accuracy and scheduling overhead. We implemented DIBS in the Click Modular Software Router. Extensive experiments have demonstrated the feasibility and effectiveness of DIBS in achieving short-term bandwidth differentiation. Compared with the representative PDD algorithm waiting time priority, DIBS can achieve better bandwidth differentiation when the inter-class packet size distributions are different. Compared with the representative weighted fair queueing algorithm per-packet generalized processor sharing, DIBS can achieve more accurate or comparable bandwidth differentiation at various workload situations, with better delay differentiation and lower cost. Furthermore, we apply DIBS to an adaptive intrusion detection and response system. Results demonstrate its efficiency for the rule-based rate-limiting function.