[ns] tcp modified to AIMD and AIPD problem.....
cheng pong lo
n2689134@dec4000.cc.ncku.edu.tw
Sun Aug 18 05:16:40 2002
This is a multi-part message in MIME format.
------=_NextPart_000_0026_01C246C6.C42F2C50
Content-Type: text/plain;
charset="big5"
Content-Transfer-Encoding: quoted-printable
hi, every seniors:
i have some problems about NS simulation=20
=20
i try to simulate the aipd algorithm, i do not know what should be=20
done to NS to simulate AIPD.
(`a comparison of end-to-end congestion control algorithms : the =
case=20
of AIMD and AIPD`, kang-wond lee,tae-eun kim,vaduvur bharghavan, =20
infocom 2001)
so i try to modify reno first. i let both ssthresh and cwnd become=20
cwnd*(1 - beta*lossrate ), ( i try to estimate lossrate by sequence=20
number) when three duplicate acks received every time.=20
the result shows that, some tcp-aipd`s sending rate (cwnd/srtt)
behave as semicircle. and there are many rectangles on this =
semicurcle.
i try to modify cwnd only, and keep ssthresh as aimd (0.5).
but the result is the same.
=20
have any senior do research about modifying congestion window(cwnd) =
when duplicate ack is received?
could you tell me the detail modification please ,=20
thank you
------=_NextPart_000_0026_01C246C6.C42F2C50
Content-Type: text/html;
charset="big5"
Content-Transfer-Encoding: quoted-printable
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META http-equiv=3DContent-Type content=3D"text/html; charset=3Dbig5">
<META content=3D"MSHTML 6.00.2600.0" name=3DGENERATOR>
<STYLE></STYLE>
</HEAD>
<BODY bgColor=3D#ffffff>
<DIV><FONT size=3D4>hi, every seniors:</FONT></DIV>
<DIV><FONT size=3D4></FONT> </DIV>
<DIV><FONT size=3D4> i have some problems about =
NS=20
simulation </FONT></DIV>
<DIV>
<DIV align=3Djustify><FONT =
size=3D4> </FONT></DIV>
<DIV align=3Djustify><FONT size=3D4> i try to simulate =
the aipd=20
algorithm, i do not know what should be </FONT></DIV>
<DIV align=3Djustify><FONT size=3D4> done to NS to =
simulate=20
AIPD.</FONT></DIV>
<DIV align=3Djustify><FONT size=3D4> (`a comparison of =
end-to-end=20
congestion control algorithms : the case </FONT></DIV>
<DIV align=3Djustify><FONT size=3D4> of AIMD and =
AIPD`, kang-wond=20
lee,tae-eun kim,vaduvur bharghavan, </FONT></DIV>
<DIV align=3Djustify><FONT size=3D4> infocom =
2001)</FONT></DIV>
<DIV align=3Djustify><FONT size=3D4></FONT> </DIV>
<DIV align=3Djustify><FONT size=3D4> so i try to =
modify reno=20
first. i let both ssthresh and cwnd become </FONT></DIV>
<DIV align=3Djustify><FONT size=3D4> cwnd*(1 - =
beta*lossrate ),=20
( i try to estimate lossrate by sequence </FONT></DIV>
<DIV align=3Djustify><FONT size=3D4> number) when =
three duplicate=20
acks received every time. </FONT></DIV>
<DIV align=3Djustify><FONT size=3D4> the result shows =
that, some=20
tcp-aipd`s sending rate (cwnd/srtt)</FONT></DIV>
<DIV align=3Djustify><FONT size=3D4> behave as =
semicircle. and=20
there are many rectangles on this semicurcle.</FONT></DIV>
<DIV align=3Djustify><FONT size=3D4></FONT> </DIV>
<DIV align=3Djustify><FONT size=3D4> i try to modify =
cwnd only,=20
and keep ssthresh as aimd (0.5).</FONT></DIV>
<DIV align=3Djustify><FONT size=3D4> but the result is =
the=20
same.</FONT></DIV>
<DIV align=3Djustify><FONT size=3D4> </FONT></DIV>
<DIV align=3Djustify><FONT size=3D4> have any senior =
do research=20
about modifying congestion window(cwnd) when duplicate ack is=20
received?</FONT></DIV>
<DIV align=3Djustify><FONT size=3D4> could you tell me =
the detail=20
modification please , </FONT></DIV>
<DIV align=3Djustify><FONT size=3D4></FONT> </DIV>
<DIV align=3Djustify><FONT size=3D4> =
=20
=
=20
=
=20
thank =
you</FONT></DIV></DIV></BODY></HTML>
------=_NextPart_000_0026_01C246C6.C42F2C50--