[e2e] Two questions on simulation of AQM

[Zhang Miao]

Hi, all
   Recently I did some simulation with single congested gateway in ns-2. 
I have two questions related to the simulation details. I would be grateful 
if anyone could provide some advices or references.

(1) Where should the action(Marking/Dropping) be taken?
    In the code of RED, action is taken in enque(), ie., at the input of the
queue. From intuition, better performance can be got when action is taken in
deque(), ie., at the output of the queue. The reason is that it can shorten
the feedback latency, especially when the RTT is similar to the queueing delay.
What I want to know is whether there is any implemention limitation in the router.

(2) What policy should be used in the reverse direction?
    When we test the performance of one AQM scheme, we only pay attention to the 
scheme applied in one direction. I just observed that the policy in the reverse 
diretion also has some influence to the result. ACK packets are controlled by it.

    Thanks a lot!


*****************************************************************
*    Zhang Miao                                                 *
*    Ph.D candidate,Department of Computer Science & Technology * 
*    Tsinghua University,Beijing,China(100084)                  *
*    Tel: (8610)-62785822                                       *
*    Email: zm at csnet1.cs.tsinghua.edu.cn                        *
*    Web: http://netlab.cs.tsinghua.edu.cn/~zm (domestic only ) *
*****************************************************************