[ns] Tracegraph installation Problem under Windows
Hello everybody: I installed tracegraph under windows according to this instruction: Installation instructions on Windows 1)Download tracegraph and mglinstaller from the page make sure it is windows version http://diament.ists.pwr.wroc.pl/~tracegr/tgform.html 2)unzip tracegraph202.zip to say, C:\ 3)execuate mglinstaller.exe, when it asks for directory name, enter C:\tracegraph202 4) add C:\tracegraph202\bin\win32 to your enivronment PATH 5)how to add to environment PATH?? right click on Mycomputer and select properties From properties select Advanced tab Select environment path variables Select PATH and add your path to it Thats it tracegraph is now ready!! This is the procedure i followed while installing tracegraph on both windows and linux (if you dont have matlab installed on your machine.) but I get this error : The application has failed to start because libmmfile.dll was not found. Re-installing the application may fix the problemwhen I want to run trgraph. I need to add, I already have matlab7 installed and it's path is : C:\Program Files\MATLAB704\bin\win32. I hope somebody can help me with error. Thank you, Sara _ Got something to buy, sell or swap? Try Windows Live Expo ttp://clk.atdmt.com/MSN/go/msnnkwex001001msn/direct/01/?href=http://expo.live.com/
[ns] energy-model in ns-2.29: sendDown
Hi All, I have a question regarding the energy model and wireless-phy in ns-2.29. In the function WirelessPhy::sendDown(Packet *p) we see the following code: v1.23 http://nsnam.cvs.sourceforge.net/nsnam/ns-2/mac/wireless-phy.cc?revision=1.23view=markup 207 if (em()) { 208 //node is off here... 209 if (Is_node_on() != true ) { 210 Packet::free(p); 211 return; 212 } 213 if(Is_node_on() == true Is_sleeping() == true){ 214 em()- DecrSleepEnergy(NOW-update_energy_time_, 215 P_sleep_); 216 update_energy_time_ = NOW; 217 218 } v1.22 http://nsnam.cvs.sourceforge.net/nsnam/ns-2/mac/wireless-phy.cc?revision=1.22view=markup 190 if (em()) 191 if ((em()-node_on() != true) || (em()-sleep())) { 192 Packet::free(p); 193 return; 194 } It seems rational to me that we should also free the packet if the node is on and sleeping (as in v1.22). According to v 1.23-v1.25, the wireless-phy can send a packet down, even if it's sleeping? What is the rationale? Best regards, Pieter -- Pieter De Mil Department of Information Technology (INTEC) Ghent University - IBBT e: [EMAIL PROTECTED] a: Gaston Crommenlaan 8 bus 201 - 9050 Ghent - Belgium t: +32 9 3314981
[ns] Wireless God
I'm trying to use the God::instance()-IsReachable(node1,node2) method and the simulation stops saying God is off. How can I turn God on? ggeorgak
Re: [ns] Wireless God
Ok, don't bother: set $god [create-god 2] $god on ggeorgak On Wednesday 23 August 2006 13:07, Giorgis Georgakoudis wrote: I'm trying to use the God::instance()-IsReachable(node1,node2) method and the simulation stops saying God is off. How can I turn God on? ggeorgak
Re: [ns] What's wrong with this code?
who is agents_(source) ? Why you have twice the same comand? $ns connect $client_tcp $server_tcp $ns connect $server_tcp $client_tcp On 8/22/06, 임인택 (Intaek Lim) [EMAIL PROTECTED] wrote: Hello, I just wrote simple tcl script for ns but it did not work. - set ns [new Simulator] set client [$ns node] set server [$ns node] $ns duplex-link $client $server 10Mb 10ms DropTail $ns duplex-link-op $client $server queuePos 0.5 set client_tcp [new Agent/TCP/FullTcp] set server_tcp [new Agent/TCP/FullTcp] lappend agents_(source) $client_tcp $server_tcp $server_tcp listen $ns attach-agent $client $client_tcp $ns attach-agent $server $server_tcp $ns connect $client_tcp $server_tcp $ns connect $server_tcp $client_tcp set client_app [new Application] $client_app attach-agent $client_tcp set server_app [new Application] $server_app attach-agent $server_tcp $client_app start $server_app start $client_app send 100 _ When I run this script, I get an error like below: --- Classfier::no-slot{} default handler (tcl/lib/ns-lib.tcl) --- _o12: no target for slot -1 _o12 type: Classifier/Hash/Dest content dump: classifier _o12 0 offset 0 shift 2147483647 mask 1 slots slot 0: _o30 (Classifier/Port) -1 default -- Finished standard no-slot{} default handler -- Please tell me what should I do. thanks. -- Renata Vidal Triste é o destino de quem tenta vencer as batalhas e ter sucesso nos ataques sem cultivar o espírito da iniciativa (A Arte da Guerra – SUN TSU – pg 102)
[ns] integration of WLAN UMTS
hello it's the second time i send you this message because since 12 Aug i didn't got any response for my request so pls if any body can help i really need it So i'm sending back my request : Hi everyone, I was reading so many articles of the integration WLAN/UMTS which include some ns simulation results and I have been searching for ns code but I didnt found any so if anyone has codes for such integration I would like to know it will be very helpful. So Im waiting for anybody help, Thank you in advance. _ Check the weather nationwide with MSN Search: Try it now! http://search.msn.com/results.aspx?q=weatherFORM=WLMTAG
[ns] ns ftp
From what I see there is no ftp implementation in ns, more precisely what seems like ftp application is a tcl Application/FTP object that the only thing it does is to send packets through a tcp connection. So is there a real ftp implementation that complies to the standard? ggeorgak
Re: [ns] ns ftp
On Wednesday 23 August 2006 16:21, Pedro Fortuna wrote: No, not that I know. But it shouldnt be too hard to extend the existing one... OK you are definitely kidding, have you seen the rfc specification of FTP? At least for me hard is not defined by how mind challenging is a task but by the time it consumes, and my estimate is that it needs considerable time to have an FTP implementation. Especially since we are talking on an abandoned project like ns. I am considering moving my simulations to PlanetLab and waiting for the arrival of Internet2 but I am hugely disappointed by the potential not exploited of ns. At least when I see an Application/FTP interface I am expecting to do what it says. If I hadn't checked the source code deviating from my main work I wouldn't know ( and it's bad not to deliver what you promiss). ggeorgak On 8/23/06, Giorgis Georgakoudis [EMAIL PROTECTED] wrote: From what I see there is no ftp implementation in ns, more precisely what seems like ftp application is a tcl Application/FTP object that the only thing it does is to send packets through a tcp connection. So is there a real ftp implementation that complies to the standard? ggeorgak
Re: [ns] ns ftp
I am sorry for the aggresiveness of my mail. The free implementation is too unreliable for what I am trying to do as I want very accurate results. The use of emulation is possible but still there are better ways to do it (PlanetLab). As for the open-source spirit you are completely right. I apologize again, it was in the spur of the moment. ggeorgak On Wednesday 23 August 2006 16:56, you wrote: You have to understand that not everyone has the same requirements you have. Very few ppl would need a complete and rigorous implementation of FTP in NS. Many don't even care if it is FTP at all... they just want to drive a continuous flow of tcp traffic through the simulated scenario. When you do some simulations, you do not need to implement everything, you just need to figure out what is important for your studies. There's two things that come to mind that you could probably do. 1) adapt a free implementation of FTP to a new Application/FTP source and sink 2) use real FTP traffic (NS2 can be made to bridge between two real life FTP applications). I'm not saying that those are easy to accomplish, but they are certainly feasible. As for the unmaintainability of NS, I share your concerns, but there's nothing we can do except to contribute. That's the spirit of open-source. In the meanwhile, NS3 is being planed and will be available some day. Best Regards, Pedro Fortuna INESC Porto
Re: [ns] ns ftp
You have to understand that not everyone has the same requirements you have. Very few ppl would need a complete and rigorous implementation of FTP in NS. Many don't even care if it is FTP at all... they just want to drive a continuous flow of tcp traffic through the simulated scenario. When you do some simulations, you do not need to implement everything, you just need to figure out what is important for your studies. There's two things that come to mind that you could probably do. 1) adapt a free implementation of FTP to a new Application/FTP source and sink 2) use real FTP traffic (NS2 can be made to bridge between two real life FTP applications). I'm not saying that those are easy to accomplish, but they are certainly feasible. As for the unmaintainability of NS, I share your concerns, but there's nothing we can do except to contribute. That's the spirit of open-source. In the meanwhile, NS3 is being planed and will be available some day. Best Regards, Pedro Fortuna INESC Porto On 8/23/06, Giorgis Georgakoudis [EMAIL PROTECTED] wrote: On Wednesday 23 August 2006 16:21, Pedro Fortuna wrote: No, not that I know. But it shouldnt be too hard to extend the existing one... OK you are definitely kidding, have you seen the rfc specification of FTP? At least for me hard is not defined by how mind challenging is a task but by the time it consumes, and my estimate is that it needs considerable time to have an FTP implementation. Especially since we are talking on an abandoned project like ns. I am considering moving my simulations to PlanetLab and waiting for the arrival of Internet2 but I am hugely disappointed by the potential not exploited of ns. At least when I see an Application/FTP interface I am expecting to do what it says. If I hadn't checked the source code deviating from my main work I wouldn't know ( and it's bad not to deliver what you promiss). ggeorgak On 8/23/06, Giorgis Georgakoudis [EMAIL PROTECTED] wrote: From what I see there is no ftp implementation in ns, more precisely what seems like ftp application is a tcl Application/FTP object that the only thing it does is to send packets through a tcp connection. So is there a real ftp implementation that complies to the standard? ggeorgak -- Cumprimentos, Pedro Fortuna
Re: [ns] Diffusion trace format help needed
Hello in a message like this: r 6.317259333 _1_ RTR --- 1486717614 diffusion 0 [13a 1 2 800] --- [2:255 1:255 32 1] r means a message was received, it could be a s for a message sent. the number next to it is the time _1_ means the node 1 received the message RTR is the type of message, I don't know exactly what it means... the next number is a identification of the message, each message has an unique number. If two lines appear with the same number, it means it is the same message (as an example, a message may appear 2 times, when it was sent and when it was received) I don't know what the diffusion 0 means and also don't know what the final numbers means.. Anyone can clarify the pending stuff? On 8/22/06, Chitra Magar [EMAIL PROTECTED] wrote: Hi everyone, I have been carrying out ns2 simulation using diffusion protocol for wireless sensors. I am get the following traces s 0.01000 _0_ AGT --- 30504276 diffusion 0 [0.00 0 0] --- [0:254 0:255 32 0] r 0.01000 _0_ RTR --- 30504276 diffusion 0 [0.00 0 0] --- [0:254 0:255 32 0] s 0.01000 _1_ AGT --- 503803425 diffusion 0 [0.00 0 0] --- [1:254 1:255 32 0] r 0.01000 _1_ RTR --- 503803425 diffusion 0 [0.00 0 0] --- [1:254 1:255 32 0] s 0.01000 _2_ AGT --- 1374534831 diffusion 0 [0.00 0 0] --- [2:254 2:255 32 0] r 0.01000 _2_ RTR --- 1374534831 diffusion 0 [0.00 0 0] --- [2:254 2:255 32 0] s 0.01000 _3_ AGT --- 1733080622 diffusion 0 [0.00 0 0] --- [3:254 3:255 32 0] r 0.01000 _3_ RTR --- 1733080622 diffusion 0 [0.00 0 0] --- [3:254 3:255 32 0] s 0.01000 _4_ AGT --- 1486717601 diffusion 0 [0.00 0 0] --- [4:254 4:255 32 0] r 0.01000 _4_ RTR --- 1486717601 diffusion 0 [0.00 0 0] --- [4:254 4:255 32 0] s 0.01000 _5_ AGT --- 1454595779 diffusion 0 [0.00 0 0] --- [5:254 5:255 32 0] r 0.01000 _5_ RTR --- 1454595779 diffusion 0 [0.00 0 0] --- [5:254 5:255 32 0] s 0.01000 _6_ AGT --- 402363792 diffusion 0 [0.00 0 0] --- [6:254 6:255 32 0] r 0.01000 _6_ RTR --- 402363792 diffusion 0 [0.00 0 0] --- [6:254 6:255 32 0] s 0.01000 _7_ AGT --- 23210729 diffusion 0 [0.00 0 0] --- [7:254 7:255 32 0] r 0.01000 _7_ RTR --- 23210729 diffusion 0 [0.00 0 0] --- [7:254 7:255 32 0] s 0.01000 _8_ AGT --- 1500389329 diffusion 0 [0.00 0 0] --- [8:254 8:255 32 0] r 0.01000 _8_ RTR --- 1500389329 diffusion 0 [0.00 0 0] --- [8:254 8:255 32 0] Diffusion 0 does it mean no data sent ? I am just confused because I have seen trace format like TCP 100 which means 100 bytes transferred. I want to calculate the packets send and packets received using diffusion protocol. I am getting the diffusion 0 at AGT layer. I would be grateful if you could give me some idea how to calculate the total data sent and total data recieved by a particular node. looking forward for your reply. Yours sincerly, Chitra Magar East China University of Science and Technology Shanghai China
Re: [ns] measuring energy - smac
take a look at the Chapter 19 of the NS Manual: Energy Model in ns On 8/22/06, Lambrakis Emmanouil [EMAIL PROTECTED] wrote: hi ns users i want to measure the energy benefits of the s-mac protocol comparing with 802.11.For this reason i want to calculate the energy consuption of a wireless network using awk. Could you help me with the calculation of the energy consuption Thanks !!!
[ns] make error
Hello I'm trying make and I get the following error. I am using ns-allinone-2.29 on cygwin. --- $ make g++ -c -Wall -DTCP_DELAY_BIND_ALL -DNO_TK -DTCLCL_CLASSINSTVAR -DNDEBUG -DUSE_ SHM -DHAVE_LIBTCLCL -DHAVE_TCLCL_H -DHAVE_LIBOTCL1_11 -DHAVE_OTCL_H -DHAVE_LIBTK 8_4 -DHAVE_TK_H -DHAVE_LIBTCL8_4 -DHAVE_TCL_H -DHAVE_CONFIG_H -DNS_DIFFUSION -D SMAC_NO_SYNC -DCPP_NAMESPACE=std -DUSE_SINGLE_ADDRESS_SPACE -Drng_test -I. -I/ns -allinone-2.29/tclcl-1.17 -I/ns-allinone-2.29/otcl-1.11 -I/ns-allinone-2.29/incl ude -I/ns-allinone-2.29/include -I/usr/include/pcap -I./tcp -I./sctp -I./common -I./link -I./queue -I./adc -I./apps -I./mac -I./mobile -I./trace -I./routing -I. /tools -I./classifier -I./mcast -I./diffusion3/lib/main -I./diffusion3/lib -I./d iffusion3/lib/nr -I./diffusion3/ns -I./diffusion3/filter_core -I./asim/ -I./qs - I./diffserv -I./satellite -I./wpan -o emulate/ether.o emulate/ether.cc emulate/ether.cc:47:26: net/ethernet.h: No such file or directory emulate/ether.cc: In static member function `static u_char* Ethernet::nametoaddr (const char*)': emulate/ether.cc:107: error: 'struct ifreq' has no member named 'ifr_data' make: *** [emulate/ether.o] Error 1 Really appreciate your help. Yehia __ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com
[ns] [bug] having segmentation fault when running scenario more than 1 sec. on AOMDV
[Bug Report] - Category: Run Time - Segmentation Fault Package: ns 2.1b4a OS:linux Environment Variables: do not know - Description: Description of Problem: i installed AOMDV and i used ./setdest to genenrate a scenario file and cbrgen to generate a pattern traffic file. but when i run the tcl file it works and loads the scenario and then gives a segmentation fault message. the out.tr is generated but it has data tell he starts to send data packets.
[ns] how to send only 2 tcp packets
can anyone please tell me how to send only 2 tcp packets in a simulation. these 2 pkts must also be rxmitted if nessecary. thanks in advance
Re: [ns] I have a problem in simulation .... while the compilation is okay
I reckon a new event ID is assigned every time you schedule an event. You can create a new packet and have it sent down in Mac. Kind Regards, Mo On Wed, 23 Aug 2006 04:28:26 +1000, Bilal AbuQadous [EMAIL PROTECTED] wrote: First of all; thank you (*Mo Li* )for your response. As you see in my code; that i create new packet in [void Mac::recv(Packet* p, Handler* h)], which is sent down by using sendDown() function. after debugging: the problem is stopped inside sendDown(): if(!abstract_) s.schedule(hRes_, intr_, txt); when I tracked the code inside the scheduler.cc; the progressing stops in: Scheduler::schedule(Handler* h, Event* e, double delay) . . . . if (e-uid_ 0) { printf(Scheduler: Event UID not valid!\n\n); abort(); } . . Thus, can you tell me how can I create new packet inside MAC layer with new valid Event UID. your help is highly appreciated Kind Regards On 8/22/06, Mo Li [EMAIL PROTECTED] wrote: Bilal AbuQadous wrote: Dear ns-users; I use ns-allinone-2.29 under cygwin/winXP. I faced the following problem after changing the MAC in ns2.29. during the simulation I got the follwoing Error: Scheduler: Event UID not valid! the code is here ( I made the changes just in the recv() function): void Mac::recv(Packet* p, Handler* h) { printf( \n Now you are in Mac layer); int i; int k; int r; int errcount; int recR; //===start=== BiLaL if (hdr_cmn::access(p)-direction() == hdr_cmn::UP) { if (hdr_mac::access(p)- ftype() == MF_FEC_R) { Red_= hdr_mac::access(p)-Red(); drop(p); printf( \n Redundancy is CHANGED); return; } if (hdr_cmn::access(p)-size()%framesize_ != 0) k=1; else k=0; k +=(hdr_cmn::access(p)-size()+ Red_)/framesize_ ; for (i=0;i= k; i++) { r=(rand()%100); if (r = p_) errcount++; } if (errcount Red_) { drop(p);//Packet::free(p); //discard the received packet printf(\n one packet has been lost); Packet *p = Packet::alloc(); hdr_cmn* ch = HDR_CMN(p); hdr_mac* mhr = HDR_MAC(p); mhr-ftype()=MF_FEC_R; ch-uid() = 0; ch-ptype() = PT_MAC; ch-iface() = -2; ch-error() = 0; ch-size()=64; ch-direction() = hdr_cmn::DOWN; // *** changing the Redundancy strategy mhr- Red()= errcount; callback_=h; mhr-set(MF_FEC_R, index_); state(MAC_SEND); printf(\n telling the sender by new R); sendDown(p); } else { // the Redundant Data must be removed hdr_mac* mhr = HDR_MAC(p); recR = mhr- Red(); hdr_cmn::access(p)-size()-= recR; //===END=== BiLaL sendUp(p); } return; } callback_ = h; hdr_mac* mh = HDR_MAC(p); mh-set(MF_DATA, index_); state(MAC_SEND); //===FEC (Redundancy) BiLaL printf(\n Here we adding the R to TCP packet); hdr_cmn* ch = HDR_CMN(p); ch-size() += Red_; // sendDown(p); the problem stop inside the sendDown() function and the problem as I told you first: Scheduler: Event UID not valid! I look forward to your help as soon as possible. thank you in advance Kind Regards One possible reason to the Event UID not valid! problem is that the event has been scheduled twice. You can print the relevant event in the scheduler.cc, and see if that helps. Kind Regards, Mo
Re: [ns] Help with wireless simulation
no extension is available for ns-2.29 u have to use ns-2.27 if u want to simulate sensor networks using ns-2,but it has lots of probs woth eenrgy model ,energy model only works for mac 802_11 protocol . Lambrakis Emmanouil wrote: Hi ns-users, I am new to this group. I would like some help with wireless simulations. I want to calculate the energy end delay for wireless topology. There are same scripts in the ns_ users but all for wired networks. Is there anyone have any awk script to calculate delay ( or a web page I could visit ) from an old format wireless .tr file like this s 1.0 _0_ AGT --- 0 cbr 512 [ 0.00 0 0] [energy 1000.00] --- [0:0 1:0 32 0] [0] 0 0 r 1.0 _0_ RTR --- 0 cbr 512 [0.00 0 0] [energy 1000.00] --- [0:0 1:0 32 0] [0] 0 0 s 1.0 _0_ RTR --- 0 cbr 512 [0.00 0 0] [energy 1000.00] --- [0:0 1:0 32 0] [0] 0 0 One more question. I want to install the nrlns2Extender-nightlybuilt to simulate sensor networks. Does anyone know if the extension PHENOM is available for ns2.29 version ? Sorry for my English Best Regards! your sincere Manolis -- View this message in context: http://www.nabble.com/Help-with-wireless-simulation-tf1966782.html#a5957585 Sent from the ns-users forum at Nabble.com.