Re: [ns] Simulating node failures
Actually it seems like we can use error models for this, I guess there should be some type of extension from generic error models to support specific node errors. Dhammika On 5/3/06, Hai Nam <[EMAIL PROTECTED]> wrote: > On 5/3/06, Dhammika Pathirana <[EMAIL PROTECTED]> wrote: > > > > Hi, > > > > What is the easiest way to simulate node failures (in wireless nets) > > in ns-2. I was trying to find a way to simulate node failures but I > > could not find any built in mechanisms for that. It seem like node > > failures are simulated by individual link failures. But then they are > > not the same. > > > > thanks > > > > Dhammika > > > > > > Hi, > > I'm working on a simulation of wireless networks, too. I was thinking > about it before, but I've never read up on it. I think a simple (but > not pretty) way to simulate a failed mobile node is to move it far > away from the others. > -- > Chào thân ái, With kind regards, Cordialement, > > Hai-Nam NGUYEN > http://www.jcisio.com >
[ns] Question about CSMA in 802.11
HI friends, I have a question about the behavior of Physical Carrier sensing in NS2. Can any one help me to clarify it? According to my understanding, in the ns-2 code, physical carrier sensing is handled by a timer. When there is a packet with a received power higher than CST the timer is set to the end of that packet's transmission. Before the timer expires, the station assumes the channel is busy and can not start a transmission. My question is that: If there comes a new packet with a received power higher than CST before the timer expires, will ns-2 extend the timer of channel busy to the end of the second packet's transmission? or It just assumes the channel is idle when the previous timer expires? Many Thanks, Hui Ma University of Washington _ 享用世界上最大的电子邮件系统― MSN Hotmail。 http://www.hotmail.com
Re: [ns] Usual Mistake
Not bytes received - it is always an integer. x/y should be typecasted to float to get correct answer. use 4.0, 5.0 as someone else had replied. > > Hi, > thanks for your response, > I dont know Exactly but i think bytes recieved is > integer. > > Regards > Hamid Regards > > --- "Sita S. Krishnakumar" <[EMAIL PROTECTED]> > wrote: > >> Both formulas are the same and should give the same >> result. >> The difference you see in your example cases are due >> to the fact that all >> operations are integer. If you change it so as to do >> a float calculation, >> either formula will give the same result. >> Regards, >> Sita >> >> >> > >> > Hi All, >> > Most people used below instruction for >> calculating a >> > bandwidth >> > #Calculate the bandwidth (in MBit/s) and write it >> to >> > the files >> > puts $grfd "$now [expr $bw1/$time*8/1024/1024] >> [expr >> > $bw2/$time*8/1024/1024] [expr >> $bw3/$time*8/1024/1024]" >> > >> > But I Think The Correct one is : >> > #Calculate the bandwidth (in MBit/s) and write it >> to >> > the files >> > puts $grfd "$now [expr $bw1*8/$time/1024/1024] >> [expr >> > $bw2/$time*8/1024/1024] [expr >> $bw3/$time*8/1024/1024]" >> > >> > I did this on ns >> > set x 5; >> > set y 2; >> > puts " [expr $x/$y*8/2/2]"; >> > output :4 >> > but >> > set x 5; >> > set y 2; >> > puts " [expr $x*8/$y/2/2]"; >> > output: 5 >> > >> > Whay is your ideas? >> > >> > Best Regards >> > Hamid Reza >> > >> > >> > __ >> > Do You Yahoo!? >> > Tired of spam? Yahoo! Mail has the best spam >> protection around >> > http://mail.yahoo.com >> > >> > >> >> >> > > > __ > Do You Yahoo!? > Tired of spam? Yahoo! Mail has the best spam protection around > http://mail.yahoo.com >
[ns] Relation Between CSThresh_, RXThresh_ , Distance, Pt_ and Reciever Sensivity in 802.11b
Hi All, I have Static WLAN 802.11b Scenario. I used TwoRayGround For Propagation. I want to Calculate Good Positioning of My Nodes. In Spec Tec of Real WLAN 802.11b Card we Have Receiver Sensivity And Power Output. I knew that Power output Equal with Pt_. But How can I relate Between other parameters. Can Anyone Help me in this issue? Best Regards Hamid Reza __ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com
Re: [ns] Usual Mistake
Both formulas are the same and should give the same result. The difference you see in your example cases are due to the fact that all operations are integer. If you change it so as to do a float calculation, either formula will give the same result. Regards, Sita > > Hi All, > Most people used below instruction for calculating a > bandwidth > #Calculate the bandwidth (in MBit/s) and write it to > the files > puts $grfd "$now [expr $bw1/$time*8/1024/1024] [expr > $bw2/$time*8/1024/1024] [expr $bw3/$time*8/1024/1024]" > > But I Think The Correct one is : > #Calculate the bandwidth (in MBit/s) and write it to > the files > puts $grfd "$now [expr $bw1*8/$time/1024/1024] [expr > $bw2/$time*8/1024/1024] [expr $bw3/$time*8/1024/1024]" > > I did this on ns > set x 5; > set y 2; > puts " [expr $x/$y*8/2/2]"; > output :4 > but > set x 5; > set y 2; > puts " [expr $x*8/$y/2/2]"; > output: 5 > > Whay is your ideas? > > Best Regards > Hamid Reza > > > __ > Do You Yahoo!? > Tired of spam? Yahoo! Mail has the best spam protection around > http://mail.yahoo.com > >
[ns] Otcl syntax problem - accessing properties on objects
Hi all, Just a quick newbie question to do with Otcl syntax - how do I access properties on objects? for example in NS-2, wireless nodes $nodes are of type 'MobileNode'. They have an ragent_ property on them (found out by running the info Tcl command). So now how do I access ragent_ from within the TCL script, I've tried $node ragent_ but this doesn't seem to work. Even trying to access '$node random-motion' doesn't seem to work, even though the command 'set $node random-motion 1' seems to work fine for setting the property - but how about accessing it? Any help would be greatly appreciated! Many Thanks, Paul - Paul Kiddie 1st Year PhD Electronic & Electrical Engineering School of Electronic, Electrical & Computer Engineering, University Of Birmingham, Edgbaston, Birmingham, B15 2TT
[ns] providing more realistic nodes mobility traces referred to ad hoc network simulation
Hi all, I write for asking your help. I have an excited idea about providing more realistic nodes mobility traces referred to ad hoc network simulation, such as in NS2, and I plan to implement it in this summer vacation. I will describe my general idea and the reason why I need your help as follows. It's well known that several synthetic mobility models have been proposed for (or used in) the performance evaluation of ad hoc network protocols, such as Random Waypoint Mobility Model, Gauss-Markov Mobility Model and Reference Point Group Mobility Mobel[1]. It's a good way to introduce realistic traces into simulations, just like the Internet traffic archive. The problem is we don't have abundant trace archives since the ad hoc network has not been used widely. My idea is that, we could utilize the cyber-games entities traces in ad hoc simulation. There are a kind of game in which people control hundreds of entities to fight with each other, and when the game is over, people could save the replays in files and share them. Is it exciting if there are hundreds to thousands entities moving on a map under real players' decision? I believe it's a perfect scenario to evaluate the protocol performance of the Mobile Ad Hoc Network, MANET. I want to parse the replay of the game “StarCraft”, and export the entities traces into ns-format. Then the traces could be used in performance evaluation of MANET protocol, and could be helpful in developing more accurate mobility models. Many worthwhile works could be done after the first step, but I don't think it's a good idea to talk too much without doing anything. Google is holding its Summer-Of-Code 2006 to support open source project[2]. Before applying for their financial support on my idea, I need find a proper mentor first. *I wonder is there anyone who could be my project mentor and give me appropriate guide to implement this idea.* It will not cost your too much time. In this movement, students are required to write project proposal and develop individually, while mentors should give advice about what to do, and check up whether students have finished the work after the movement this September. Besides, the mentor will be paid 500$ for their work, and I will commit all my codes to NS community. If ( you agree that my idea is worthwhile && could guide me) Please response me, and be my mentor; else if ( you agree that my idea is worthwhile && cannot guide me) Please give me some advice, and recommend someone else who could be my mentor if possible; else Just disregard this mail, and I am sorry to bother you :); It is worth noting that the applying process will be end in 8th, May. In other words, the 8th, May is my deadline of finding a proper mentor. As a mentor, one should be experienced in reasearch of wireless/ad hoc network, and(or) development of ns2. Any advice, help and recommend will be appricated deeply. Look forward to hearing from you. With my best regards, -- Luo Gang Reference [1] Tracy Camp, Jeff Boleng, Vanessa Davies, A Survey of Mobility Models for Ad Hoc Network Research, Wireless Communication & Mobile Computing (WCMC): Special issue on Mobile Ad Hoc Networking: Research, Trends and Applications, vol. 2, no. 5, pp. 483-502, 2002. [2] http://code.google.com/summerofcode.html
Re: [ns] the neighbors in networks simulator,no one can help me?
Hello, god.cc is used for wireless networks simulation in general. I don't know if IsNeighbor is used, but you can search for "IsNeighbor" in ~ns to verify that (grep -r ...). About AODV (and any wireless routing protocols that use "HELLO message", I thinks), look at ~ns/aodv/aodv.cc, in the function recvHello(), you'll see that a neighbor is a node from whom you can hear directly (HELLO message won't be forwarded). If you want the "correct" definition of "neighbor" or other terminologies, take a look at RFC 3753. In my opinion, you should not redefine this term as you said in a few messages ago. It'd better invent a different term. Finally, about the mailing lists. Usually you'll have a reply in less than 24h, but it might take a few days. You can do some research, or post in another mailing list (as I can see uptill now, 2 long replies from MANET WG, so I try to be short ^_^). Good luck (& bon courage *_* ). On 5/3/06, www triste <[EMAIL PROTECTED]> wrote: > > Hello everybody, > i ask this question 3 times,but no one help me > i delate the Isneighbor fonction in god.cc and i recompile the > ns-allinone,nothing change,no probleme when i compile my tcl file!! > that s mean that aodv package don't use the Isneighbor fonction > help me please,really i can't see solution,where can i find the detection > of neighbor in aodv code.please help > Thanks lot, > -- Chào thân ái, With kind regards, Cordialement, Hai-Nam NGUYEN http://www.jcisio.com
Re: [ns] TCL Simulation script/C++ interaction query wrt AODV routing protocol
Hi, Many thanks for your speedy reply, the 'bind' command is exactly what I was thinking of doing for tying C++ variables with TCL variables, but now I'm wondering where I should put these bind commands (as far as which source file is concerned) to be able to access it as: '$node set isMalicous_ 0' within my simulation TCL script. If I were to put it directly into the AODV::AODV constructor within AODV.CC as specified in the NS documentation for an Agent, how would I access the individual node's isMalicious_ property? I am currently using the node-config API to instantiate all my ad hoc nodes, should I be doing this to have to kind of granular control I need? Many Thanks, Paul - Paul Kiddie 1st Year PhD Electronic & Electrical Engineering School of Electronic, Electrical & Computer Engineering, University Of Birmingham, Edgbaston, Birmingham, B15 2TT -Original Message- From: debojit boro [mailto:[EMAIL PROTECTED] Sent: 04 May 2006 12:12 To: [EMAIL PROTECTED] Subject: Re: [ns] FW: TCL Simulation script/C++ interaction query wrt AODV routing protocol Hi, I'm doin project on the same thing but slightly different from what u r doing.I think to access the variables both in Tcl scripts as well as C++ code u need to bind the variables such as : "bind("isMalicious_",&isMalicious_)".I think u better consult the NS manual.Plzz reply me if it works. Debojit Boro. M.Tech(IT) 4th Sem Deptt. of Computer Science and Information Technology Tezpur University, Napaam,Tezpur Dist-Sonitpur,Assam,India --- Paul Kiddie <[EMAIL PROTECTED]> wrote: > > > Hi all, > > > > I'm new to NS-2 simulation and in my research im > looking at MANET > > routing protocols, namely the AODV routing > protocol. In my simulation > > work I would like to simulate 'malicious' nodes on > the network, but be > > able to control whether a node is malicious or not > via a TCL script, > > and also in what ways it is malicous. > > > > For example, it would be nice in my TCL script to > enable/disable > > malicousness by: $ns at $x "$node($i) set > isMalicous_ 0". I need to be > > able to access these variables both in my > simulation TCL script, and > > also in the AODV C++ script, to take the desired > action. Can anyone > > shed any light on how I would go about this? > > > > Many Thanks, > > Paul > > > > - > > > > Paul Kiddie > > 1st Year PhD Electronic & Electrical Engineering School of > > Electronic, Electrical & Computer > Engineering, > > University Of Birmingham, > > Edgbaston, > > Birmingham, > > B15 2TT > > > __ Yahoo! India Answers: Share what you know. Learn something new. http://in.answers.yahoo.com
Re: [ns] Usual Mistake
Hello Hamid! I think the first expression would be correct if $bw1 was expressed in bits, not octets(8 bits/8=1 Byte, not 8 bits*8=1 byte). But since the packets' sizes are measured in bytes(octets), i think that 8 is not necessary anymore, the measuring unit will result in MBps without multipying with 8. Regards, Alexandra hamid reza <[EMAIL PROTECTED]> wrote: Hi All, Most people used below instruction for calculating a bandwidth #Calculate the bandwidth (in MBit/s) and write it to the files puts $grfd "$now [expr $bw1/$time*8/1024/1024] [expr $bw2/$time*8/1024/1024] [expr $bw3/$time*8/1024/1024]" But I Think The Correct one is : #Calculate the bandwidth (in MBit/s) and write it to the files puts $grfd "$now [expr $bw1*8/$time/1024/1024] [expr $bw2/$time*8/1024/1024] [expr $bw3/$time*8/1024/1024]" I did this on ns set x 5; set y 2; puts " [expr $x/$y*8/2/2]"; output :4 but set x 5; set y 2; puts " [expr $x*8/$y/2/2]"; output: 5 Whay is your ideas? Best Regards Hamid Reza __ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com - How low will we go? Check out Yahoo! Messengers low PC-to-Phone call rates.
[ns] Usual Mistake
Thanks For you Responses, But I dont know Are Bytes Recieved From Agents Store as double in "bw" variable or store at int in it? if store as double anything be ture. but if store as int this bug exist in our codes. Regards Hamid Reza __ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com
Re: [ns] Usual Mistake
I think it's just because the results are rounded to integers after each step. Replace set x 5 by set x 5.0 will return the correct answer. With kind regards, -- Hai-Nam NGUYEN http://www.jcisio.com
Re: [ns] How can I estimate delay in Diffserv
Hi First, if you want to estimate the delay, you need AWK or perl script to analyse the trace file. Second, there is no such a thing as a class delay. You can estimate mean delay of packets belonging to some class (or the maximum delay), but you still need to analyse the trace file. Sincerely, Alexander Sayenko (PhD) Assistant Telecommunication laboratory, MIT department University of Jyvaskyla, Finland > > > Dear Ns-users: > I have a question about Ns 2.27: > I am training in Diffserv. > How can I estimate delay for each class? > Send me a TCL file that estimated delay please. > > yours, > ghazale khojaste > > > - > Talk is cheap. Use Yahoo! Messenger to make PC-to-Phone calls. Great rates starting at 1¢/min.
[ns] Warning message
Hi I created a new type of header for my agent and of course I put a new offset that I called off_source_ but when I run my tcl script I obtain this warning message warning: no class variable Agent/Source::off_source_ How can I correct that. Thanx __ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com
[ns] FW: TCL Simulation script/C++ interaction query wrt AODV routing protocol
> Hi all, > > I'm new to NS-2 simulation and in my research im looking at MANET > routing protocols, namely the AODV routing protocol. In my simulation > work I would like to simulate 'malicious' nodes on the network, but be > able to control whether a node is malicious or not via a TCL script, > and also in what ways it is malicous. > > For example, it would be nice in my TCL script to enable/disable > malicousness by: $ns at $x "$node($i) set isMalicous_ 0". I need to be > able to access these variables both in my simulation TCL script, and > also in the AODV C++ script, to take the desired action. Can anyone > shed any light on how I would go about this? > > Many Thanks, > Paul > > - > > Paul Kiddie > 1st Year PhD Electronic & Electrical Engineering > School of Electronic, Electrical & Computer Engineering, > University Of Birmingham, > Edgbaston, > Birmingham, > B15 2TT >
[ns] How can I estimate delay in Diffserv
Dear Ns-users: I have a question about Ns 2.27: I am training in Diffserv. How can I estimate delay for each class? Send me a TCL file that estimated delay please. yours, ghazale khojaste - Talk is cheap. Use Yahoo! Messenger to make PC-to-Phone calls. Great rates starting at 1¢/min.
[ns] MANET > Having DIFFERENT Routing Protocol in ONE Network
Hi everyone, (In my project, I've focused on "DSR" and Black hole attacks ...) I wanna implement different routing protocol in my testbed : DSR and modified DSR I will be appreciate you if I recieve some hints to treat this query: How can I FLAG a node uses MALICIOUS "dsragent.cc" or a NORMAL "dsragent.cc"? (where the both dsragent should be compiled simultaneously) thanx in advance
[ns] Usual Mistake
Hi All, Most people used below instruction for calculating a bandwidth #Calculate the bandwidth (in MBit/s) and write it to the files puts $grfd "$now [expr $bw1/$time*8/1024/1024] [expr $bw2/$time*8/1024/1024] [expr $bw3/$time*8/1024/1024]" But I Think The Correct one is : #Calculate the bandwidth (in MBit/s) and write it to the files puts $grfd "$now [expr $bw1*8/$time/1024/1024] [expr $bw2/$time*8/1024/1024] [expr $bw3/$time*8/1024/1024]" I did this on ns set x 5; set y 2; puts " [expr $x/$y*8/2/2]"; output :4 but set x 5; set y 2; puts " [expr $x*8/$y/2/2]"; output: 5 Whay is your ideas? Best Regards Hamid Reza __ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com
Re: [ns] Connecting agents--many to many
Hi, You can do that by specifying a port number for your agent and then attach your agent to the nodes through that port. That way, an agent can communicate with any other agent by specifying its address and port id. Example : - In .h and .cc files: #define CLUSTERING_PORT 113 struct hdr_ip *ih = HDR_IP(p); ihpk->daddr() = dest; // destination node address ihpk->saddr() = index; ihpk->sport() = CLUSTERING_PORT; ihpk->dport() = CLUSTERING_PORT; - In .tcl script : #creating clustering agents and attaching them to nodes through port 113. for {set j 0} {$j < $opt(nn) } {incr j} { set clagent_($j) [new Agent/Clustering [$node_($j) node-addr] ] $node_($j) attach $clagent_($j) 113 } Hope it helps. Regards. On 5/2/06, [EMAIL PROTECTED] <[EMAIL PROTECTED]> wrote: > > I am currently trying to develop a simulation in which a wireless sensor > network with sleeping nodes has an ad-hoc "backbone" of nodes with a higher > duty cycle to reduce latency. In order for this scenario to work, however, > nodes must be able to negotiate with logical children and neighboring nodes > in order to nominate backbone nodes. I have implemented this logic as an > Agent subclass; however, I am now realizing that Agents seem to have only a > one-to-one relationship. Is there a workaround for this (so that an Agent > may have multiple agents it can communicate with?) I have tried calls > through the OTCL linkage directly to the interpreter to try and reconnect > nodes on the fly, but this does not seem to work. Does anyone have any > suggestions? > >
[ns] moving mobile nodes
Hi all, I m trying to implement a simple DSR protocol. Whenever i create a mobile node such as "set mobilenode [new Node/MobileNode]" the nodes don't tend to move but if created as "set node [$ns_ node]" the mobile nodes moves.Why is it happening??Can anyone please plzz tel me the reason.Plzz help me put. Thanks in advance. Regards Debojit Boro. __ Yahoo! India Answers: Share what you know. Learn something new. http://in.answers.yahoo.com
Re: [ns] port number
Hi, These tcl commands are implemented in Node.cc or MobileNode.cc in the command( ) method which is used to add tcl commands. Bye. On 4/28/06, Chams D <[EMAIL PROTECTED]> wrote: > > Hi, > can anyone tell me how to use the procedure $node id > or $node port ??? > and also where are theses methodes implemented, > because I need to implement a new method like this. > thanx > > __ > Do You Yahoo!? > Tired of spam? Yahoo! Mail has the best spam protection around > http://mail.yahoo.com > >