evaluate TCP Low Priority congestion control algorithm on ns3

 A point to point channel is built to test the performance of vegas. The link bandwidth is 3Mbps, propagation delay is 50 ms, maxmium buffer length(100ms). Three flows are involved in tests.

When three flows all take ledbat for rate control.

 The sending rate got from sender side:
在這裏插入圖片描述
 packets receiving rate got form receiver(calculated at every five seconds):
在這裏插入圖片描述
 one way delay of received packets:
在這裏插入圖片描述
 Intra-protocol bandwidth allocation fairness can be guaranteed.

When two tcp-lp flows share links with a reno flow

 The flow3 takes reno for rate control.
 The sending rate got from sender side:
在這裏插入圖片描述
 packets receiving rate got form receiver:
在這裏插入圖片描述
 Comments: Oh no, it seems low priority tcp flows show no sign of inferior rate to tcp reno flow.
[1] TCP-LP: A Distributed Algorithm for Low Priority Data Transfer
[2] 低優先級的TCP擁塞控制TCP-LP

發表評論
所有評論
還沒有人評論,想成為第一個評論的人麼? 請在上方評論欄輸入並且點擊發布.
相關文章