Long Le

The Effects of Active Queue Management on Tcp Application Performance: an Experimental Performance Evaluation Long Le

Active queue management (AQM) has been proposed by networking researchers and the Internet Engineering Task Force as a measure to preserve and improve Internet performance but has not been thoroughly evaluated under realistic conditions. This book investigates the effects of several AQM algorithms on the performance of TCP applications under realistic conditions. Our primary results are that many existing AQM algorithms do not perform as well as expected when they are used with packet dropping. However, when combined with packet marking, AQM algorithms significantly improve network and application performance over conventional drop-tail queues. Moreover, AQM enables network operators to run their networks near saturation levels with only modest increases in average response times. If packet marking is unavailable, the book presents a new form of differential treatment of flows that can be used with packet dropping and achieves a similar positive performance improvement. The book also develops a new AQM algorithm that can balance between loss rate and queuing delay to improve the overall system performance.