Exercise 4: ECN

Task is to study how use of Explicit Congestion Notification will affect network traffic.

Your answer

Using ns2 simulator, you should set up proper simulation cases to answer following questons with our simulation results (include relevant data for your answer).

Does use of ECN improve network throughput?
Does use of ECN improve individual flow throuhput?
If only part of flows use ECN, does throughput of non-ECN flows degenerate?
Are there less lost packets in network?
Is network utilization better?

Return is due to 2001-12-10T1545 to G2 wing course locker.

Help for simulator setup

You can use network topology used in first exercise as basis for simulation. It is not the best setup for this study so you are encouraged to have more complicated setup.

Following changes for example.tcl are needed to study ECN:

Change queue management from dsRED/edge to normal RED

< $ns simplex-link $e1 $e2 10Mb 5ms dsRED/edge
< $ns simplex-link $e2 $e1 10Mb 5ms dsRED/edge
---
> $ns simplex-link $e1 $e2 10Mb 5ms RED
> $ns simplex-link $e2 $e1 10Mb 5ms RED

Remove all DS-related settings from queues (setNumPrec, addPolicyEntry...)

Activate ECN on routers

$qE1E2 set setbit_ 1
$qE2E1 set setbit_ 1

Activate ECN on (selected) sources
```
set tcp3 [new Agent/TCP/Sack1]
set sink3 [new Agent/TCPSink]
$tcp3 set ecn_ 1
```
It would be interesting also compare Tahoe TCP (Agent/TCP) without ECN to Reno with SACK (Agent/TCP/Sack1) and ECN.

Tietoverkkolaboratorio on nyt osa Tietoliikenne- ja tietoverkkotekniikan laitosta. Tällä sivulla oleva tieto voi olla vanhentunutta.

Kurssien ajantasainen tieto on MyCourses-palvelussa.

Tämän sivun sisällöstä vastaavat ja Webmaster.
Sivua on viimeksi päivitetty 20.11.2001 17:42.
URI: http://www.netlab.tkk.fi/opetus/s38180/s01/slides/harjoitus4.shtml
[ TKK > Sähkö- ja tietoliikennetekniikan osasto > Tietoverkkolaboratorio > Opetus ]

Kysy

Anna palautetta!