Some questions about the simulations done on ns2 about DCCP

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Dear all,

I am a beginner of DCCP. At the moment I am writing my undergraduate thesis about DCCP.
I tried some simulations on ns2 with DCCP module by Nils-Erik Mattsson. There are some questions that I hope I could get some helps.

1. I am focusing on DCCP CCID2. The TCP-Like Congestion Control. And the DCCP TCP-Like Agent is referenced from draft-ietf-dccp-ccid2-04. And up to now, the latest profile for CCID2 is draft-ietf-dccp-ccid2-10 (since March 2005). And I tried to do some simulations, I found out that the Sack TCP(Also Newreno TCP) is always more aggressive than DCCP in many conditions. And I notices the changes log noted that the version 05 of the profile is changed according to Nils-Erik Mattsson's feedback.
Finally in version 10, the byte-counting is disabled for making DCCP more aggressive.

So is it an expected result that the Sack/Newreno TCP is more aggressive than DCCP? DCCP is unreliable transfer protocol, so DCCP should have better throughput than TCP.

I have read a ref paper: Performance Evaluations of DCCP for Bursty Traffic
The paper concludes a different result shows that Sack TCP is less aggressive than DCCP in some condition.

2. Is there a updated DCCP -CCID2's module in ns2?

If someone is interested in my work, I am happy to put some plots and scripts on the net.

Thanks in advanced,

Tai-hua Ma
Student of  IT Faculty
Macao University of Science and Technology
EMAIL: 0209853ai011033@xxxxxxxxxxxxxxxxxxx

[Index of Archives]     [Linux Kernel Development]     [Linux DCCP]     [IETF Annouce]     [Linux Networking]     [Git]     [Security]     [Linux Assembly]     [Bugtraq]     [Yosemite]     [MIPS Linux]     [ARM Linux]     [Linux Security]     [Linux RAID]     [Linux SCSI]     [DDR & Rambus]

  Powered by Linux