A Parallel Virtual Queue Structure for Active Queue Management

The Adaptive RED proposed by Feng et al . [5] is shown to have small packet delay and queue length variation for long-life TCP traffic such as FTP connection with a large file size. However, a great portion of Internet traffic is shortlife web and UDP traffic. Most web traffic has a small file size and its TCP session is mainly operated in the slow start phase with a small congestion window size. Since the file size is small, dropping short-life TCP (and UDP) packets is not very effective in alleviating congestion level at a bottleneck router. From the viewpoint of TCP, one or several packet losses in its slow start phase lead to extra delay for retransmission and even cause TCP timeout. This delay severely degrades the performance of delivering short messages such as web pages and web browsers experience a long waiting time even with a high speed network. We first show that the Adaptive RED is vulnerable to these shortlife TCP traffic and propose a virtual parallel queue structure as a new active queue management scheme (AQM). The idea is to separate the long-life and short-life (including UDP) traffic into two different virtual queues. The first queue is to run the droptail policy and work for the short-life TCP and UDP packets. In order to have a small mean delay, the service rate of this drop-tail queue is dynamically determined by its virtual queue length. The remaining long-life traffic is directed to an Adaptive RED virtual queue. Even the available bandwidth is shared with the drop-tail queue, the simulation results show that the queue length variation of the RED queue is still located in a desired region. Note that both virtual queues share the same physical buffer memory. Those packets in the drop-tail queue will not be dropped unless the shared buffer is overflowed. This parallel virtual queue structure not only keeps the benefits of RED such as high utilization and small delay, but also greatly reduces the packet loss rate at the router.