ns3 + Evalvid

2,366 views
Skip to first unread message

Juliana Freitag Borin

unread,
Jan 30, 2012, 6:27:16 AM1/30/12
to ns-3-...@googlegroups.com
Hi all,

have any of you used Evalvid
(http://140.116.72.80/~smallko/ns2/Evalvid_in_NS2.htm) with ns-3? I'm
planning to integrate both of them, so any help would be greatly
appreciated.


Best regards,
Juliana.

--
Juliana Freitag Borin, PhD
Computer Networks Laboratory
Institute of Computing - UNICAMP
PO 6176 - Av. Albert Einstein, 1251 - Cidade Universitaria
Zip Code 13084-971 - Campinas/SP - Brazil
fax: +55 (19) 3521-0336
Email: jul...@ic.unicamp.br

billy Pinheiro

unread,
Feb 9, 2012, 9:25:25 AM2/9/12
to ns-3-...@googlegroups.com
Hi Juliana,

I did that some months ago.

Its work good and is not so complicated to do.

[]s

2012/1/30 Juliana Freitag Borin <jul...@ic.unicamp.br>
--
You received this message because you are subscribed to the Google Groups "ns-3-users" group.
To post to this group, send email to ns-3-...@googlegroups.com.
To unsubscribe from this group, send email to ns-3-users+...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/ns-3-users?hl=en.


Shengyang Chen

unread,
Feb 24, 2012, 9:29:04 AM2/24/12
to ns-3-users
Hi Billy:

I was wondering if you could kindly provide a working example with
EvalVid in NS3 here. I'm stuck in a problem with it at the moment.

I've introduced EvalVid model into my own wireless topology
implementation. However when I was running the simulation trying
analysing the throughput, I got very low throughput result which was
close to 0. I thought there might be something wrong with my topology
so I went back to run the default example provided by EvalVid (the
evalvid-client-server.cc file). This time I found that although the
default packet size was set to 1024 byte and the default max packet
number was 100 in the example itself, the result showed that only 13
bytes where received at the client side, which looks quite strange to
me.

I was wondering if you have met similar problems with EvalVid in NS3
before and if your could provide me some hints with this problem. Any
help would be appreciated.

Regards,

Shengyang

On 2月9日, 下午2时25分, billy Pinheiro <haquiti...@gmail.com> wrote:
> Hi Juliana,
>
> I did that some months ago.
>
> Its work good and is not so complicated to do.
>
> []s
>
> 2012/1/30 Juliana Freitag Borin <juli...@ic.unicamp.br>
>
>
>
> > Hi all,
>
> > have any of you used Evalvid
> > (http://140.116.72.80/~smallko/ns2/Evalvid_in_NS2.htm) with ns-3? I'm
> > planning to integrate both of them, so any help would be greatly
> > appreciated.
>
> > Best regards,
> > Juliana.
>
> > --
> > Juliana Freitag Borin, PhD
> > Computer Networks Laboratory
> > Institute of Computing - UNICAMP
> > PO 6176 - Av. Albert Einstein, 1251 - Cidade Universitaria
> > Zip Code 13084-971 - Campinas/SP - Brazil
> > fax: +55 (19) 3521-0336
> > Email: juli...@ic.unicamp.br
>
> > --
> > You received this message because you are subscribed to the Google Groups
> > "ns-3-users" group.
> > To post to this group, send email to ns-3-...@googlegroups.com.
> > To unsubscribe from this group, send email to
> > ns-3-users+...@googlegroups.com.
> > For more options, visit this group at
> >http://groups.google.com/group/ns-3-users?hl=en.- 隐藏被引用文字 -
>
> - 显示引用的文字 -

billy Pinheiro

unread,
Mar 6, 2012, 12:18:48 PM3/6/12
to ns-3-...@googlegroups.com
Hi Shengyang,

Sry for the later response :)

The default max packet(100) actually does nothing. there is, but i did not use.

For me the module works well with the example(evalvid-client-server.c)

please, post your code(not in the mail body) and i will see :)

Regards,

Billy Pinheiro.

2012/2/24 Shengyang Chen <home5...@gmail.com>

Shengyang Chen

unread,
Mar 6, 2012, 2:52:10 PM3/6/12
to ns-3-users
Hi Billy:

The code I was using was the exact example provided in EvalVid model,
with only the PhyRxOK/PhyRxError statement detecting scheme added to
detect packet receiving so that throughput and loss could be
calculated.

Or maybe you could provide me some ideas of measuring throughput and
loss in a better way? Thanks a lot.

Regards,

Shengyang

On 3月6日, 下午5时18分, billy Pinheiro <haquiti...@gmail.com> wrote:
> Hi Shengyang,
>
> Sry for the later response :)
>
> The default max packet(100) actually does nothing. there is, but i did not
> use.
>
> For me the module works well with the example(evalvid-client-server.c)
>
> please, post your code(not in the mail body) and i will see :)
>
> Regards,
>
> Billy Pinheiro.
>
> 2012/2/24 Shengyang Chen <home555...@gmail.com>
> > > >http://groups.google.com/group/ns-3-users?hl=en.-隐藏被引用文字 -

Anders Branderud

unread,
Mar 11, 2012, 7:50:32 AM3/11/12
to ns-3-...@googlegroups.com
Hello Shengyang!
I am interested in using EvalVid+ with NS3. I would be very thankful if you (or anyone else reading this) could you send me your implementation of EvalVid+ and NS3?

Furthermore I am interested if anyone has any suggestions of how to combine EvalVid+ with adding FEC-code-packets (preferably LDPC, available e.g. at www.openfec.org).

Thanks in advance!!

-- Kind regards, Anders Branderud

[Personal blog] Will of the Creator  : Logical reasons - based on scientific premises - for the existence of a Super intelligent and Orderly Creator and that He hasn't left His sapient creatures without an Instruction Manual - Torah ['books of Moses'] - to ascertain, and aspire to, His purpose.

[Company] Anders Branderud IT Solutions - www.abitsolutions.org
> > > > ns-3-users+unsubscribe@googlegroups.com.
> > > > For more options, visit this group at
> > > >http://groups.google.com/group/ns-3-users?hl=en.-隐藏被引用文字 -
>
> > > - 显示引用的文字 -
>
> > --
> > You received this message because you are subscribed to the Google Groups
> > "ns-3-users" group.
> > To post to this group, send email to ns-3-...@googlegroups.com.
> > To unsubscribe from this group, send email to
> > ns-3-users+unsubscribe@googlegroups.com.

Shengyang Chen

unread,
Mar 27, 2012, 7:30:38 AM3/27/12
to ns-3-users
Hi all:

I have recently downloaded a new version of EvalVid model from Gercom
UFPA, which I think is designed for ns3 version 3.13. There are some
modifications for the bugs in the previous version and they seem to be
fine.

However when I was trying running the example it provided in ns 3.13,
it terminated with a signal SIGSEGV runtime error which indicated
there were behaviors of accessing invalid memory locations. Yet when I
was trying debugging with gdb, I couldn't find any places that related
to problems like null pointers in the backtrace. Meanwhile during the
trace process of the example, I noticed that the input file stream in
the evalvid-server.cc didn't actually write anything in the while loop
in the setup() function (line 190 in evalvid-server.cc). The values of
id, type, size etc. were all zero, which seems to be unreasonable.

The only change I've made in the example is the attribute value
"SendTraceFileName" of the EvalVidServerHelper in line 93. I replaced
the file name "st_highway_cif.st" with my own test video file name,
which I believe is the correct way to use the model for our own
simulation.

Could anyone give me a hint related to this problem please? If any
further details are needed please note as well. Thanks.

Regards,

Shengyang

Shengyang Chen

unread,
Mar 27, 2012, 7:34:32 AM3/27/12
to ns-3-users
Hi Anders:

Sorry for the late reply. As I mentioned in my previous reply I was
just using the exact example evalvid-client-server.cc provided in the
model package downloaded from Gercom when I bumped into the issue. So
I guess that's the implementation you need?

Still, I would be glad to continue discussion with you about the NS3 +
EvalVid simulation in further research.

Regards,

Shengyang

On 3月11日, 下午12时50分, Anders Branderud <anders.brande...@gmail.com>
wrote:
> Hello Shengyang!
> I am interested in using EvalVid+ with NS3. I would be very thankful if you
> (or anyone else reading this) could you send me your implementation of
> EvalVid+ and NS3?
>
> Furthermore I am interested if anyone has any suggestions of how to combine
> EvalVid+ with adding FEC-code-packets (preferably LDPC, available e.g. atwww.openfec.org).
>
> Thanks in advance!!
>
> -- *Kind regards, Anders Branderud*
>
> [Personal blog]  <goog_1576108163>Will of the Creator<http://www.proofexistencegod.com/will-of-god/will-of-god/>  : Logical
> > > > > > ns-3-users+...@googlegroups.com.
> > > > > > For more options, visit this group at
> > > > > >http://groups.google.com/group/ns-3-users?hl=en.-隐藏被引用文字 -
>
> > > > > - 显示引用的文字 -
>
> > > > --
> > > > You received this message because you are subscribed to the Google
> > Groups
> > > > "ns-3-users" group.
> > > > To post to this group, send email to ns-3-...@googlegroups.com.
> > > > To unsubscribe from this group, send email to
> > > > ns-3-users+...@googlegroups.com.

Shengyang Chen

unread,
Mar 30, 2012, 6:39:48 AM3/30/12
to ns-3-users
Hi Billy:

Regarding to my last post of the SIGNSEV problem when I was trying the
example evalvid-client-server.cc, I think I have (kind of) found what
was going on there, but I'm still confused about the reason.

By tracing the process with Gdb, I noticed that when it came to around
the end of the simulation, there was one time of invoking
GetNextFrame() in the ProcessStreamData() function in evalvid-
server.cc (line 253), where the "trec_time" variable had a negative
value. This caused the "vtime" variable below it to be negative as
well, which made the first parameter (that was "vtime") of the
Schedule() event function negative, and thus caused the NS_ASSERT
(tAbsolute.IsPositive ()) to be failed, which was the place the
SIGNSEV occurred.

I don't understand at all why the program got a negative time frame
from the trace file. I was wondering if it's because there's something
wrong with the way I encode/decode the sample video. Could you maybe
give me some advices on that aspect?

The following steps are how I get the trace file used in the example
program:

1. Download sample video (lossless h264 format) from
http://www2.tkn.tu-berlin.de/research/evalvid/cif.html . I chose the
highway_cif.264 video.

2. Decode the video in yuv format in Cygwin: ffmpeg -i highway_cif.264
highway_cif.yuv

3. Create compressed raw video in Cygwin: ffmpeg -s cif -r 30 -b 64000
-bt 3200 -g 30 -i highway_cif.yuv -vcodec mpeg4 highway_cif.m4v

4. Create ISO MP4 video containing the sample frames and a hint track
which describes how to packetize the frames for the transport with
RTP:
MP4Box -hint -mtu 1024 -fps 30 -add highway_cif.m4v
highway_cif.mp4

5. Use MP4 to record the tracefile sender: MP4.exe –send 10.1.1.2 4000
1024 highway_cif.mp4 > highway_cif.st

Shengyang Chen

unread,
Mar 30, 2012, 7:03:53 AM3/30/12
to ns-3-users
Hi Billy:

Here's the content of my trace file generated. Notice the last line,
which missed the frame type and size. I was thinking maybe that's the
reason why it caused a negative trec_time value, since it was read
from the trace file line by line but this line missed 2 variables and
the value of trec_time failed to be read.

0 H 5795 6 segm at 34 ms
1 I 9697 10 segm at 68 ms
2 P 4467 5 segm at 100 ms
3 P 1765 2 segm at 134 ms
4 P 1251 2 segm at 166 ms
5 P 1262 2 segm at 200 ms
6 P 1212 2 segm at 233 ms
7 P 1078 2 segm at 265 ms
8 P 1227 2 segm at 298 ms
9 P 1264 2 segm at 331 ms
10 P 1315 2 segm at 364 ms
11 P 1196 2 segm at 397 ms
12 P 1326 2 segm at 430 ms
13 I 16101 16 segm at 464 ms
14 P 1513 2 segm at 496 ms
15 P 2150 3 segm at 529 ms
16 P 1358 2 segm at 562 ms
17 H 1675 2 segm at 595 ms
18 P 1935 2 segm at 628 ms
19 P 1328 2 segm at 661 ms
20 P 1661 2 segm at 694 ms
21 P 1631 2 segm at 727 ms
22 P 1517 2 segm at 760 ms
23 P 1405 2 segm at 793 ms
24 P 1127 2 segm at 826 ms
25 P 1061 2 segm at 859 ms
26 I 11826 12 segm at 893 ms
27 P 505 1 segm at 925 ms
28 P 820 1 segm at 958 ms
29 P 1083 2 segm at 991 ms
30 P 1351 2 segm at 1024 ms
31 P 848 1 segm at 1057 ms
32 P 1117 2 segm at 1090 ms
33 H 1083 2 segm at 1123 ms
34 P 1019 1 segm at 1156 ms
35 P 916 1 segm at 1189 ms
36 P 854 1 segm at 1222 ms
37 P 835 1 segm at 1255 ms
38 P 971 1 segm at 1288 ms
39 I 9650 10 segm at 1321 ms
40 P 735 1 segm at 1354 ms
41 P 927 1 segm at 1387 ms
42 P 873 1 segm at 1420 ms
43 P 886 1 segm at 1453 ms
44 P 721 1 segm at 1486 ms
45 P 647 1 segm at 1519 ms
46 P 583 1 segm at 1552 ms
47 P 1253 2 segm at 1585 ms
48 P 538 1 segm at 1618 ms
49 H 775 1 segm at 1651 ms
50 P 670 1 segm at 1684 ms
51 P 1352 2 segm at 1717 ms
52 I 11611 12 segm at 1751 ms
53 P 238 1 segm at 1783 ms
54 P 360 1 segm at 1816 ms
55 P 314 1 segm at 1849 ms
56 P 944 1 segm at 1882 ms
57 P 617 1 segm at 1915 ms
58 P 695 1 segm at 1948 ms
59 P 714 1 segm at 1981 ms
60 P 816 1 segm at 2014 ms
61 P 544 1 segm at 2047 ms
62 P 807 1 segm at 2080 ms
63 P 887 1 segm at 2113 ms
64 I 9527 10 segm at 2147 ms
65 H 1055 2 segm at 2179 ms
66 P 519 1 segm at 2212 ms
67 P 989 1 segm at 2245 ms
68 P 506 1 segm at 2278 ms
69 P 593 1 segm at 2311 ms
70 P 1122 2 segm at 2344 ms
71 P 679 1 segm at 2377 ms
72 P 761 1 segm at 2410 ms
73 P 911 1 segm at 2443 ms
74 P 978 1 segm at 2476 ms
75 P 1116 2 segm at 2509 ms
76 P 792 1 segm at 2542 ms
77 I 8282 9 segm at 2575 ms
78 P 569 1 segm at 2608 ms
79 P 743 1 segm at 2641 ms
80 P 603 1 segm at 2674 ms
81 H 747 1 segm at 2707 ms
82 P 747 1 segm at 2740 ms
83 P 969 1 segm at 2773 ms
84 P 734 1 segm at 2806 ms
85 P 825 1 segm at 2839 ms
86 P 837 1 segm at 2872 ms
87 P 710 1 segm at 2905 ms
88 P 565 1 segm at 2938 ms
89 P 1004 1 segm at 2971 ms
90 I 9521 10 segm at 3005 ms
91 P 489 1 segm at 3037 ms
92 P 664 1 segm at 3070 ms
93 P 959 1 segm at 3103 ms
94 P 538 1 segm at 3136 ms
95 P 640 1 segm at 3169 ms
96 P 1019 1 segm at 3202 ms
97 H 719 1 segm at 3235 ms
98 P 514 1 segm at 3268 ms
99 P 1025 2 segm at 3301 ms
100 P 504 1 segm at 3334 ms
101 P 664 1 segm at 3367 ms
102 P 760 1 segm at 3400 ms
103 I 9636 10 segm at 3434 ms
104 P 320 1 segm at 3466 ms
105 P 356 1 segm at 3499 ms
106 P 831 1 segm at 3532 ms
107 P 363 1 segm at 3565 ms
108 P 530 1 segm at 3598 ms
109 P 946 1 segm at 3631 ms
110 P 528 1 segm at 3664 ms
111 P 868 1 segm at 3697 ms
112 P 916 1 segm at 3730 ms
113 H 747 1 segm at 3763 ms
114 P 634 1 segm at 3796 ms
115 P 673 1 segm at 3829 ms
116 I 9501 10 segm at 3863 ms
117 P 398 1 segm at 3895 ms
118 P 633 1 segm at 3928 ms
119 P 690 1 segm at 3961 ms
120 P 738 1 segm at 3994 ms
121 P 701 1 segm at 4027 ms
122 P 695 1 segm at 4060 ms
123 P 417 1 segm at 4093 ms
124 P 654 1 segm at 4126 ms
125 P 462 1 segm at 4159 ms
126 P 405 1 segm at 4192 ms
127 P 329 1 segm at 4225 ms
128 I 11548 12 segm at 4259 ms
129 H 1027 2 segm at 4291 ms
130 P 383 1 segm at 4324 ms
131 P 605 1 segm at 4357 ms
132 P 538 1 segm at 4390 ms
133 P 507 1 segm at 4423 ms
134 P 661 1 segm at 4456 ms
135 P 408 1 segm at 4489 ms
136 P 691 1 segm at 4522 ms
137 P 364 1 segm at 4555 ms
138 P 869 1 segm at 4588 ms
139 P 471 1 segm at 4621 ms
140 P 604 1 segm at 4654 ms
141 I 9566 10 segm at 4688 ms
142 P 404 1 segm at 4720 ms
143 P 615 1 segm at 4753 ms
144 P 985 1 segm at 4786 ms
145 H 719 1 segm at 4819 ms
146 P 799 1 segm at 4852 ms
147 P 903 1 segm at 4885 ms
148 P 566 1 segm at 4918 ms
149 P 723 1 segm at 4951 ms
150 P 681 1 segm at 4984 ms
151 P 611 1 segm at 5017 ms
152 P 545 1 segm at 5050 ms
153 P 452 1 segm at 5083 ms
154 I 9819 10 segm at 5117 ms
155 P 419 1 segm at 5149 ms
156 P 461 1 segm at 5182 ms
157 P 504 1 segm at 5215 ms
158 P 472 1 segm at 5248 ms
159 P 529 1 segm at 5281 ms
160 P 639 1 segm at 5314 ms
161 H 719 1 segm at 5347 ms
162 P 671 1 segm at 5380 ms
163 P 721 1 segm at 5413 ms
164 P 542 1 segm at 5446 ms
165 P 645 1 segm at 5479 ms
166 P 650 1 segm at 5512 ms
167 I 9725 10 segm at 5546 ms
168 P 470 1 segm at 5578 ms
169 P 776 1 segm at 5611 ms
170 P 553 1 segm at 5644 ms
171 P 548 1 segm at 5677 ms
172 P 515 1 segm at 5710 ms
173 P 543 1 segm at 5743 ms
174 P 885 1 segm at 5776 ms
175 P 403 1 segm at 5809 ms
176 P 380 1 segm at 5842 ms
177 H 719 1 segm at 5875 ms
178 P 601 1 segm at 5908 ms
179 P 507 1 segm at 5941 ms
180 I 9670 10 segm at 5975 ms
181 P 210 1 segm at 6007 ms
182 P 543 1 segm at 6040 ms
183 P 507 1 segm at 6073 ms
184 P 270 1 segm at 6106 ms
185 P 599 1 segm at 6139 ms
186 P 381 1 segm at 6172 ms
187 P 432 1 segm at 6205 ms
188 P 614 1 segm at 6238 ms
189 P 437 1 segm at 6271 ms
190 P 560 1 segm at 6304 ms
191 P 576 1 segm at 6337 ms
192 I 9632 10 segm at 6371 ms
193 H 971 1 segm at 6403 ms
194 P 345 1 segm at 6436 ms
195 P 447 1 segm at 6469 ms
196 P 419 1 segm at 6502 ms
197 P 366 1 segm at 6535 ms
198 P 812 1 segm at 6568 ms
199 P 325 1 segm at 6601 ms
200 P 512 1 segm at 6634 ms
201 P 548 1 segm at 6667 ms
202 P 456 1 segm at 6700 ms
203 P 579 1 segm at 6733 ms
204 P 412 1 segm at 6766 ms
205 I 9606 10 segm at 6799 ms
206 P 187 1 segm at 6832 ms
207 P 470 1 segm at 6865 ms
208 P 549 1 segm at 6898 ms
209 H 719 1 segm at 6931 ms
210 P 536 1 segm at 6964 ms
211 P 333 1 segm at 6997 ms
212 P 718 1 segm at 7030 ms
213 P 360 1 segm at 7063 ms
214 P 470 1 segm at 7096 ms
215 P 512 1 segm at 7129 ms
216 P 528 1 segm at 7162 ms
217 P 487 1 segm at 7195 ms
218 I 9637 10 segm at 7228 ms
219 P 341 1 segm at 7261 ms
220 P 440 1 segm at 7294 ms
221 P 608 1 segm at 7327 ms
222 P 784 1 segm at 7360 ms
223 P 567 1 segm at 7393 ms
224 P 620 1 segm at 7426 ms
225 H 719 1 segm at 7459 ms
226 P 661 1 segm at 7492 ms
227 P 660 1 segm at 7525 ms
228 P 767 1 segm at 7558 ms
229 P 605 1 segm at 7591 ms
230 P 685 1 segm at 7624 ms
231 I 7283 8 segm at 7657 ms
232 P 436 1 segm at 7690 ms
233 P 736 1 segm at 7723 ms
234 P 493 1 segm at 7756 ms
235 P 888 1 segm at 7789 ms
236 P 461 1 segm at 7822 ms
237 P 519 1 segm at 7855 ms
238 P 801 1 segm at 7888 ms
239 P 473 1 segm at 7921 ms
240 P 512 1 segm at 7954 ms
241 H 663 1 segm at 7987 ms
242 P 806 1 segm at 8020 ms
243 P 376 1 segm at 8053 ms
244 I 8317 9 segm at 8087 ms
245 P 308 1 segm at 8119 ms
246 P 484 1 segm at 8152 ms
247 P 507 1 segm at 8185 ms
248 P 605 1 segm at 8218 ms
249 P 385 1 segm at 8251 ms
250 P 588 1 segm at 8284 ms
251 P 691 1 segm at 8317 ms
252 P 674 1 segm at 8350 ms
253 P 674 1 segm at 8383 ms
254 P 476 1 segm at 8416 ms
255 P 671 1 segm at 8449 ms
256 I 8403 9 segm at 8482 ms
257 H 915 1 segm at 8515 ms
258 P 388 1 segm at 8548 ms
259 P 612 1 segm at 8581 ms
260 P 578 1 segm at 8614 ms
261 P 546 1 segm at 8647 ms
262 P 618 1 segm at 8680 ms
263 P 624 1 segm at 8713 ms
264 P 393 1 segm at 8746 ms
265 P 732 1 segm at 8779 ms
266 P 484 1 segm at 8812 ms
267 P 535 1 segm at 8845 ms
268 P 683 1 segm at 8878 ms
269 I 8250 9 segm at 8912 ms
270 P 391 1 segm at 8944 ms
271 P 497 1 segm at 8977 ms
272 P 460 1 segm at 9010 ms
273 H 691 1 segm at 9043 ms
274 P 336 1 segm at 9076 ms
275 P 393 1 segm at 9109 ms
276 P 635 1 segm at 9142 ms
277 P 789 1 segm at 9175 ms
278 P 502 1 segm at 9208 ms
279 P 591 1 segm at 9241 ms
280 P 567 1 segm at 9274 ms
281 P 787 1 segm at 9307 ms
282 I 8133 8 segm at 9340 ms
283 P 252 1 segm at 9373 ms
284 P 368 1 segm at 9406 ms
285 P 410 1 segm at 9439 ms
286 P 454 1 segm at 9472 ms
287 P 424 1 segm at 9505 ms
288 P 774 1 segm at 9538 ms
289 H 691 1 segm at 9571 ms
290 P 299 1 segm at 9604 ms
291 P 435 1 segm at 9637 ms
292 P 472 1 segm at 9670 ms
293 P 509 1 segm at 9703 ms
294 P 569 1 segm at 9736 ms
295 I 8180 8 segm at 9769 ms
296 P 535 1 segm at 9802 ms
297 P 776 1 segm at 9835 ms
298 P 548 1 segm at 9868 ms
299 P 587 1 segm at 9901 ms
300 P 591 1 segm at 9934 ms
301 P 515 1 segm at 9967 ms
302 P 523 1 segm at 10000 ms
303 P 548 1 segm at 10033 ms
304 P 542 1 segm at 10066 ms
305 H 691 1 segm at 10099 ms
306 P 522 1 segm at 10132 ms
307 P 482 1 segm at 10165 ms
308 I 8095 8 segm at 10198 ms
309 P 376 1 segm at 10231 ms
310 P 470 1 segm at 10264 ms
311 P 527 1 segm at 10297 ms
312 P 511 1 segm at 10330 ms
313 P 444 1 segm at 10363 ms
314 P 534 1 segm at 10396 ms
315 P 648 1 segm at 10429 ms
316 P 421 1 segm at 10462 ms
317 P 579 1 segm at 10495 ms
318 P 335 1 segm at 10528 ms
319 P 528 1 segm at 10561 ms
725 1 segm at 10594 ms

Billy Pinheiro

unread,
Mar 30, 2012, 8:17:24 AM3/30/12
to ns-3-...@googlegroups.com
There is a erro in the content that u sent. The last line is wrong.

im send to u my SendTraceFile.



2012/3/30 Shengyang Chen <home5...@gmail.com>
st_highway_cif.st

Shengyang Chen

unread,
Apr 3, 2012, 4:29:40 AM4/3/12
to ns-3-users
Hi Billy:

Thanks a million for your help. The SendTraceFile of yours works fine
with the example. There must be something wrong with the steps I used
to transcode the video.

Just wondering, have you ever used EvalVid in any kind of wireless
network topology in NS3? The example used CSMA network which I think
is wired LAN, and when I tried to build a simple Wifi network with
Wifihelper, the program ended with SIGSEGV again.

All the best,

Shengyang

On 3月30日, 下午1时17分, Billy Pinheiro <haquiti...@gmail.com> wrote:
> There is a erro in the content that u sent. The last line is wrong.
>
> im send to u my SendTraceFile.
>
> 2012/3/30 Shengyang Chen <home555...@gmail.com>
> ...
>
> 阅读更多 >>
>
> st_highway_cif.st
> 51K查看下载

Preethi Chandur

unread,
Jul 28, 2012, 8:13:10 AM7/28/12
to ns-3-...@googlegroups.com
Thank you so much for posting the trace file, Billy. Thank you for the mp4 conversion procedure, Chen.

Anthony Faustine

unread,
Jul 28, 2012, 12:30:29 PM7/28/12
to ns-3-...@googlegroups.com
Help how can i run this trace conversion in linux im using fedora
> --
> You received this message because you are subscribed to the Google Groups
> "ns-3-users" group.
> To view this discussion on the web visit
> https://groups.google.com/d/msg/ns-3-users/-/fUTYKEU8TiwJ.
Message has been deleted

haikal

unread,
Nov 6, 2012, 4:38:40 AM11/6/12
to ns-3-...@googlegroups.com, jul...@ic.unicamp.br
Hello Guys,
I would like to ask about the way to run the qoe monitor module and evalvid module in ns 3 together , because after I try to run the file in each examples of each module, its work and there was a failed process after it combine together, its very urgent because thats my thesis about it , I really appreciate that someone can help me,I dont know the problem , Can it be from the library or source code?or the other reasons?thx,
here is the error picture......

Capture _Error.doc

haikal

unread,
Nov 6, 2012, 4:38:49 AM11/6/12
to ns-3-...@googlegroups.com, jul...@ic.unicamp.br
Capture _Error.doc

Md. Jalil Piran

unread,
Oct 14, 2013, 1:14:53 AM10/14/13
to ns-3-...@googlegroups.com, jul...@ic.unicamp.br
Hi All

I am new to both NS and Evalvid.

which one do you prefer to install NS2 or NS3?

i wanna work on multimedia transmission over cognitive radio networks.

for simulation, which one do you suggest?

thanks so much for your reply in advanced.

BNR INFO

unread,
Apr 16, 2014, 10:21:42 AM4/16/14
to ns-3-...@googlegroups.com, jul...@ic.unicamp.br
hello
please i'm working with evalvid for ns3
after simulation i found two fils rd and sd
so i want to decode them in order to have the reconsted video
please anybody could help 
noted that i found that ./etmp4  can make it, when can i find this tools;,,!!

Saulo da Mata

unread,
Apr 16, 2014, 12:06:29 PM4/16/14
to ns-3-...@googlegroups.com


--
You received this message because you are subscribed to the Google Groups "ns-3-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ns-3-users+...@googlegroups.com.

To post to this group, send email to ns-3-...@googlegroups.com.



--
Saulo da Mata
Federal University of Uberlândia - Brazil

Reply all
Reply to author
Forward
0 new messages