[e2e] Two questions on simulation of AQM

[Wuwei]

hi, Zhang

I think you are right if all users and routers support ECN, and as I know the 
implementation of RED in ns, a well known network simulator do support this 
function. But for not all users support ECN, Router have to drop the packets
to signal the sender. I think drop a switched packet the the top of the queue
is not an efficient way from my intuition.
Nowadays AQM is implemented at the output port or input port. I have no idea
about which one is better. Maybe it's not the problem of AQM only. We need
take the architecture of swiching and scheduling into account.
   
                

          Yours regards
            Wei WU
            wuwei at sdp.ee.tsinghua.edu.cn
|=====================================================================|
| Wei Wu, Complexity Engineering System Labs (CESL)                   |
| Electronics Engineering Department, Tsinghua University             |
| Beijing, P.R.China                                                  | 
|                                                                     | 
| Office:Room 401 Division 10,Eastern Main Building of Tsinghua Univ. |
|       (tel) 86 10 6277 3228                                         |
| Dorm  :Room 300 Apartment 14, Tsinghua Univ.                        |
|       (tel) 86 10 6277 8632                                         |
|=====================================================================|

Hi, Haining
    Perhaps you haven't understood what I mean. 
    Each marked packet should be treated as a congestion feedback signal, and it 
should reflect the "current" congestion conditions. If the packet is marked when it 
leaves the queue, it can better reflect the "current" condition, without a queueing 
delay. Otherwise, this marked packet only reflects the condition when it enters 
the queue.

>Hi,
>
>>From intuition, better performance can be got when action is taken in
>>deque(), ie., at the output of the queue.
>
>No, the packet should be dropped at the front end before any resource
>has been consumed. Also, the earlier you drop the packet, the earlier
>feedback you will get. Your intuition is wrong :) just take a
>second thought.
>



*****************************************************************
*    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 ) *
*****************************************************************