headers
Siva Reddy | 1 Feb 2005 06:27
Picon
Favicon

doubt regarding installation of DMAC

hi friends,
if any boby working on Data-gathering MAC in sensor
networks, please give me an idea about how to install
the code for DMAC given by the author in his site.i
tried a lot to install the code but i am getting lot
of errors.please help me...

thanks in advance...
siva reddy.

		
__________________________________ 
Do you Yahoo!? 
The all-new My Yahoo! - What will yours do?
http://my.yahoo.com
Permalink | Reply | 0 comments |
headers
Claudio Lavecchia | 1 Feb 2005 10:56
Picon
Picon
Favicon

How to cause and measure the hidden terminal (or collisions)?

Hello you MANET guys,

I have a challenging question for you: I am running some tests on a 
MANET testbed composed of 4 nodes equipped of Dell Truemobile 1150 WLAN 
cards and routed with OLSR.
I would like to do some performance measurements in presence of the 
hidden terminal problem. Now I believe that causing the hidden terminal 
problem with 4 nodes is almost impossible.
Is there anyone that knows how I could provoke the hidden terminal 
problem in such conditions and measure its effects?  Is there a way to 
measure collisions that happen during a 802.11 communication?

I hope my question is clear enough. Any contribution would be greatly 
appreciated.

Thx

Claudio
Permalink | Reply | 3 comments |
headers
Sherif M. ElRakabawy | 1 Feb 2005 11:20
Picon
Picon

Re: How to cause and measure the hidden terminal (orcollisions)?


Hi,

If you disable RTS/CTS you can get hidden terminals with only 3 nodes ..
You can Measure the collisions on the MAC layer by counting the number
of
unsuccessful data packet transmissions, i.e. 4 transmissions without
receiving a
MAC ACK. You may compare the values you get with a case where there are
no 
hidden terminals, i.e. RTS/CTS is enabled.

Regards

Sherif

--------------------------------------------
Sherif M. ElRakabawy
University of Dortmund
Department of Computer Science			   Tel. +49 231 755 2690
Mobile Computing Systems Group			   Fax. +49 231 755 2417
August-Schmidt-Str. 12
44227 Dortmund / Germany
WWW: http://mobicom.cs.uni-dortmund.de/~moustafa/

> -----Ursprüngliche Nachricht-----
> Von: manet-bounces <at> ietf.org [mailto:manet-bounces <at> ietf.org] 
> Im Auftrag von Claudio Lavecchia
> Gesendet: Tuesday, February 01, 2005 10:56 AM
> An: manet <at> ietf.org
(Continue reading)

Permalink | Reply | 1 comment |
headers
Imad Aad | 1 Feb 2005 11:26
Picon
Picon
Favicon

Re: How to cause and measure the hidden terminal (or collisions)?


On Tue, 1 Feb 2005, Claudio Lavecchia wrote:

> Hello you MANET guys,
>
> I have a challenging question for you: I am running some tests on a MANET 
> testbed composed of 4 nodes equipped of Dell Truemobile 1150 WLAN cards and 
> routed with OLSR.
> I would like to do some performance measurements in presence of the hidden 
> terminal problem. Now I believe that causing the hidden terminal problem with 
> 4 nodes is almost impossible.

Why?!
3 stations far enough from each other would make a good hidden node 
scenario, am I missing something? (don't forget to take the carrier sense 
threshold/range into consideration)

> Is there anyone that knows how I could provoke the hidden terminal problem in 
> such conditions and measure its effects?  Is there a way to measure 
> collisions that happen during a 802.11 communication?

On a high level: measure the throughput drop with respect to a scenario 
with no hidden stations (but the same number of stations and connections)

On a low level: use a station in monitor mode (if your card and card 
driver permit this), you'll be able to see which packets got ACKed and 
which ones did not.

I'm sure there's plenty of other ways to monitor what's happening on 
the channel.
(Continue reading)

Permalink | Reply | 0 comments |
headers
Marco | 1 Feb 2005 11:49
Picon
Favicon

Fwd: change Pt_


----------  Messaggio inoltrato  ----------

Subject: change Pt_
Date: 11:49, martedì 1 febbraio 2005
From: Marco <bragalini_m <at> tiscali.it>
To: ns-users <at> isi.edu

Hy ns-users,

I want modify Pt_  for a determined node, in C++' code.
In Mac802.11 i have made a routine which compute the distance between tx node
and rx node. For every distance, I have a corresponding value of power(pot).

In wirelessphy I have made a second routine who set Pt_:

inline double setPt(double pot){Pt_=pot; return Pt_;}

I call it in Mac802.11::send

			WirelessPhy* power;
			*
			*
			*
			pot=0.002;
			power->setPt(pot);

			*
			*
			*
(Continue reading)

Permalink | Reply | 0 comments |
headers
Claudio Lavecchia | 1 Feb 2005 15:35
Picon
Picon
Favicon

Re: How to cause and measure the hidden terminal (orcollisions)?

Thx Sherif,

Your idea of disabling the RTS is interesting and leads me further in my 
reasoning.
Let me try to explain you what I would like to achieve:

Let's imagine that I have three nodes in a row, say A,B and C. Node B 
acts as a relay for traffic between A and C. Then we have a fourth node, 
say node D.
Node D is disturbing the communication between A and C by trying to jam 
node B.
What I would like to have is node B that from time to time fails in 
forwarding packets that are directed from A to C. I tought that it would 
be possible to achieve this using node D.
If node A sends a packet to node B, but node B does not receive it 
because node D starts talking to node B at the same time, that would be 
enough for my purposes.
In this sense I do not strictly need to mount an "hidden terminal" 
scenario in the sense that for me physical separation of devices is not 
mandatory.

Now, if I understand correctly, if I apply your suggestion to the 
scenario I just described, I could have node A talking to node C through 
the relay of node B. Node D has disabled the RTS and hence it starts 
talking to node B without sensing the media.
What do I have to do to make sure that node B privileges reception of 
packets from node D (the interfering node)? Do I have to play with the 
transmission power?
How do I measure the number of MAC ACKs received? is there any statistic 
tool provided with WLAN cards?
(Continue reading)

Permalink | Reply | 0 comments |
headers
vkumar | 2 Feb 2005 05:57
Picon

question in IEEE 802.11 std

What is "post backoff"? (I have found something about it in "IEEE 802.11e
Wireless LAN for Quality of Service", Stefan Mangold etal., Proc. European
Wireless (EW 2002))
Does 802.11 standard mention something about "post-backoff"? If so where?
Also does NS2.27 implement "post-backoff"? What will happen if there is 
two backoffs (post-backoff and a pre-backoff)? i.e. If a node has already
finished post-backoff and a packet arrives into its empty queue from 
higher layer and this packet finds the channel idle, then should it do 
another backoff before transmitting the packet OR should it transmit the 
packets without any backkoff ?

Vinod Kumar
M.E. Telecommunication
IISc
Permalink | Reply | 1 comment |
headers
Shahriar Shamil Uulu | 2 Feb 2005 06:50
Picon
Favicon

Routing load


Dear All,
I have written the perl script to calculate the
normalized routing load but my results are very
strange.Can anybody tell me is it right formula for my
calculation which is given below?
Note: i am using old format of tracing 
------------------------------------------------ 
1. I am getting the receives like: if the flags are
"r" and "AGT" and "cbr" then recieves++.

2.I am getting routing packets like: if the flags are
("s" or "f") and "RTR"  and ( "aodv" or "dsr" or
"message") then routing++;

3. And i am getting nrload=routing/receives

4. I am putting the results in to the file and drawing
with xgraph, the x-axis it TIME(simulation time), and
y-axis is nrload. 
------------------------------------------------ 
Please help me to clear this doubt 
Thank you in advance ... 

=====
Shahriar Shamil Uulu 
Hyderabad Central University 
M.Tech(Computer Science)
email:shamilsons <at> yahoo.com
website: http://shamilson.narod.ru
(Continue reading)

Permalink | Reply | 0 comments |
headers
Marco | 2 Feb 2005 16:25
Picon
Favicon

Fwd: power control in mac


----------  Messaggio inoltrato  ----------

Subject: power control in mac
Date: 16:25, mercoledì 2 febbraio 2005
From: Marco <bragalini_m <at> tiscali.it>
To: ns-users <at> isi.edu

Hi ns-users,
i need a power control routine for mac802.11.
Someone can help me,please?

					Thanks, Marco

-------------------------------------------------------
Permalink | Reply | 0 comments |
headers
Yue Fang | 2 Feb 2005 17:11

Re: question in IEEE 802.11 std


Hello, Vinod:

In my understanding, there is just one backoff scheme in the standard and
Ns2.27 do imprlent it. Ppl refer it to "post backoff" as the backoff after
a packet has been transmitted in order to avoid the capture of the
channel. while "pre backoff" refers to the backoff a node execute when it
has packet to send but find channel is not idle.

For the case you described, when the node find that the backoff counter is
0 and the channel is idle for DIFS, it can transmit immediately.

Hope this helps

Regards

-----------------------------------
Yue Fang
227 Egan Research Building
Dept. of Electrical & Computer Eng.
Northeastern University
Boston, MA, 02115
(tel) 617-373-3009
-----------------------------------

On Wed, 2 Feb 2005 vkumar <at> ece.iisc.ernet.in wrote:

> What is "post backoff"? (I have found something about it in "IEEE 802.11e
> Wireless LAN for Quality of Service", Stefan Mangold etal., Proc. European
> Wireless (EW 2002))
(Continue reading)

Permalink | Reply | 0 comments |

Gmane