Mixim and 802.15.4

[Robert.C.Green]

Can anyone give me some insights on how to get the 802154Narrow
example to compile and run? I've been able to checkout and run the
802154a example.

Also, if there are any other suggestions, I'm planning on using Mixim
to study 802.15.4 in heterogeneous networks with 802.11. If anyone has
done any work in that area or has some thoughts/ideas/suggestions it
would be much appreciated.

[Jérôme Rousselot]

Hello Robert,

assuming that you downloaded and successfully compiled mixim 1.2,
try the following in a terminal:

cd mixim/examples
cd ieee802154Narrow
./runTest1.sh

Then, open the Test1.anf analysis file from within the omnet IDE
to examine the results. See the files:
mixim/examples/ieee802154Narrow/ieee802154Narrow.ned
mixim/examples/ieee802154Narrow/PhyMacHost.ned
mixim/contrib/modules/nic/Nic802154_TI_CC2420.ned

Assuming you want dual-mode hosts, you will need:
- to create your own examples directory (e.g. examples/mixedNetwork)
- copy from ieee802154Narrow the omnetpp.ini file, and the two above
mentioned ned files, rename them and change their package declaration
(edit the ned files, it is relatively simple)
- add a second Nic interface to your new Host ned file, by importing
contrib/modules/nic/Nic80211Battery.ned
- create a new network module, deriving from BaseNetwLayer, that has
additional lower gate to communicate with the additional NIC module.
To do this, you will need to create three files: a ned file, a .cc file
and a .h file. In the .cc file you will implement your algorithm that
decide on which NIC interface to use.

You will find explanations in the omnet user guide, the mixim
documentation and the mixim source code on how to do that.

If you want to study the interference between 15.4 and .11 networks,
then you will probably need to work a bit at the physical layer. mixim
provides so-called Mapping objects to model radio signals. It enables to
take into account such interference, but you may need to extend slightly
the existing Decider models.

Regards,

Jérôme

[Robert.C.Green]

Jérôme,,

Thank you so much for all of the advice. It is very helpful as I know
very little about Omnet++.

What actually happened was the following:
1 - I downloaded and compiled MiXiM
2 - There were no 802.15.4 examples in the code at all
3 - I downloaded the source from the GIT repository to get the
examples.
4 - The 802.15.4 examples had no make files

In order to build the examples I issued the following commands:

For 802.15.4Narrow:
opp_makemake -f -O out -L../../out/gcc-debug/base -L../../out/gcc-
debug/modules -L../../out/gcc-debug/tests/testUtils -lmiximbase -
lmiximmodules -I../../modules/utility -I../../base/messages -I../../
base/utils -I../../base/modules -I../../base/phyLayer -o
ieee802154Narrow

For 802.15.4a:
opp_makemake -f -O out -L../../out/gcc-debug/base -L../../out/gcc-
debug/modules -L../../out/gcc-debug/tests/testUtils -lmiximbase -
lmiximmodules -o ieee8021514a

The rest of your advice seems spectacular and I will let you know how
things work out.

Thanks so much!

[Jérôme Rousselot]

To generate the make files, try:
make -f makemakefiles

And to build:
make base modules examples

Alternatively, you can import the project into the omnet IDE and use
the Ctrl-B keyboard shortcut to build the project.

Please note that 802.15.4a concerns the UWB-IR alternative PHY layer
(500 MHz ultra wideband signals) and not the narrow band 2.4 GHz PHY
defined in 802.15.4, which is used as PHY layer for ZigBee solutions.

regards

Jérôme

[Robert.C.Green]

Also, if you would care to share some more thoughts, I am actually
studying intelligent scheduling in heterogeneous networks. In other
words, if a NIC has access to both 802.11 and 802.15.4, how does one
choose which network to use in different situations.

On Jul 12, 5:23 am, Jérôme Rousselot <jerome.rousse...@csem.ch> wrote:

[Robert.C.Green]

All good instructions. Thanks!

One question: IT seems that all the examples rely strictly on
broadcast messages. Can you point me to any examples of 802.11 or
802.15.4 where individual hosts send messages directly to each other?

[David_MIXIM]

Hello everyone,

I am also working with Mixim. Could anyone tell me how to run a new
project that is essentially a copy from the ieee802154Narrow example
in Mixim? The original example works fine, however when I copy over
the files from the example to create my own project and modify them to
avoid duplicate names, incorrect paths, etc the simulation crashes and
gives me the following error.

.gdbinit: No such file or directory.

Cannot access memory at address 0x63a1e3c

I have tried everything I could think of over the past three weeks and
got different errors but no success. I have also followed Jérôme's
instructions without any success. I have read the Omnet user manual,
the Mixim wikis, tutorials, forum posts with no success.
I am looking for fairly detailed instructions as it seems the devil is
in the details.

Any help will be immensely appreciated.

David.

[David_MIXIM]

Update to my last post, it seems the MassMobility module doesn't work
with the ieee802154Narrow example? I'm fairly new to Omnet, Mixim,
etc... but changing the mobility to ConstSpeedMobility, or
LinearMobility seemed to do the trick...

[Robert.C.Green]

Jérôme,

I made all of the changes that you suggested. Now, when I try to send
messages directly out of the new layer (DualNetworkLayer.h/cc) I get
the following message:

Error in module (DualNetworkLayer) ... send()/sendDelayed(): cannot
send message (NetwPkt)BROADCAST_MESSAGE, it is currently contained/
owned by (NetwPkt)MarkovMixed.node[1].net.BROADCAST_MESSAGE


The code that I am using is:

void DualNetworkLayer::handleUpperMsg(cMessage* msg)
{
assert(dynamic_cast<cPacket*>(msg));
NetwPkt* pkt = encapsMsg(static_cast<cPacket*>(msg));
//sendDown(pkt);
recordPacket(PassedMessage::OUTGOING,PassedMessage::LOWER_DATA,msg);
send(msg, lowerGateOut);
}

It is not very clear to me how to send messages directly to another
host.

[Karl Wessel]

I did a quick and dirty try to run the ieee802154Narrow example with
MassMobility and it worked fine. However I only tried run 9 of Config
Test1-A.
Does this problem still occur, if yes which configuration did you use?

Greetings,
Karl

On 14.07.2010 21:40, David_MIXIM wrote:
> Update to my last post, it seems the MassMobility module doesn't work
> with the ieee802154Narrow example? I'm fairly new to Omnet, Mixim,
> etc... but changing the mobility to ConstSpeedMobility, or
> LinearMobility seemed to do the trick...
>
> On Jul 13, 5:09 pm, David_MIXIM<dlaye...@gmail.com> wrote:
>
>> Hello everyone,
>>
>> I am also working with Mixim. Could anyone tell me how to run a new
>> project that is essentially a copy from the ieee802154Narrow example
>> in Mixim? The original example works fine, however when I copy over
>> the files from the example to create my own project and modify them to
>> avoid duplicate names, incorrect paths, etc the simulation crashes and
>> gives me the following error.
>>
>> .gdbinit: No such file or directory.
>>
>> Cannot access memory at address 0x63a1e3c
>>
>> I have tried everything I could think of over the past three weeks and

>> got different errors but no success. I have also followed J�r�me's

>>>>> J�r�me
>>>>>
>>
>>>>> Le vendredi 09 juillet 2010 � 16:55 +0200, Robert.C.Green a �crit :

[ayhan kiraz]

Hi dear friends. I am Ayhan KİRAZ and I am a doctora student in Sakarya University in Turkey. I study about wsn transport layer protocol in omnet mixim or mf2. I need a wsn transport layer protocol written on omnet mixim or mf2. a sample. it is very necessary for me and my thesis. Can you help me anyone? I wait for your aswer. perhaps may be any people that worked about this project.
Thank you for your interest
 
Ayhan KİRAZ
Sakarya University
mail: ayhan...@gmail.com   ayhan...@hotmail.com

[Mohd Adib Sarijari]

Hello  Robert  and All,

could you please share your code on the heterogeneous network simulation with us? I believed that you are already succeed running your simulation :) 

I would like to run the same simulation.

appreciate your kind sharing so much. have a nice day.

regards,

Adib