Exploiting fine-grained explicit router feedback towards high-performance transport protocols

摘要

The positive use of router feedback that explicitly indicates internal network conditions is a promising way to address the performance issues of congestion control in highspeed networks. In this paper, we propose SIRENS, a scalable, robust, and flexible fine-grained explicit router feedback framework where each receiver can find out the overall path status by assembling all the cumulative router feedback that indicate the status in the single hop. Such per-hop information is needed by end-hosts to design novel congestion control mechanisms or to significantly enhance the performance of traditional forms of congestion control in specific situations. As one typical application of SIRENS, we examine the optimization of TCP limited slow-start. Through the implementation and evaluation, we have made sure that a sender can shift from a slow start phase to a congestion avoidance phase without any packet drop, and can achieve more effective bandwidth utilization. We also examine multi-rate multicast congestion control as another application and evaluate the performance through simulations.