Choice of a common routing protocol for simulation

[Itiyama Sadana]

Here's what I found out

• DSR, being the oldest implemented algorithm for INET, has almost zero chances of implementation errors. But almost every feature of the protocol is programmed,hence raises the difficulty level.
• AODV is a more current version,but still implemented extensively.
• DYMO is a more recent development,hence highly simplified.                           

--

Regards,

Itiyama Sadana.

Fourth year Undergraduate,

Computer Science Engineering,

BITS,Pilani.

[Itiyama Sadana]

Could not find anything about OLSR.

Your suggestions?

[Raj Joshi]

I think older the better. We can have more reliability on the correctness of its implementation. So let us go ahead with DSR. Also DSR seems to give us more statistics.

Now the task left is the choice of Network to test DSR. I will try out all the example networks with DSR running on the top of it and come up with some conclusion.

Cheers,

--
 
 

Raj Kiran Joshi

4th Year, B.E. (Hons.) Computer Science

+91 978 500 3330 | Gmail | BITSmail

Birla Institute of Technology & Science, Pilani

Pilani Campus, Rajasthan 333 031, INDIA

[Raj Joshi]

A review of the example networks available. Will be posting as and when I complete analyzing them. So that others can take up remaining.

Fg80211:

• Few fixed hosts and others mobile
• Coordinated movement. Basically "move on self-msg" principle.
• The movement stops and then messaging starts. Some air-frames and then the fireworks. Need to figure out what exactly happens.
• But no movement during messaging. Nodes remain stationery :-|

[Raj Joshi]

Well simplicity wins over DSR. We are going for OLSR now.

Regarding networks models:

grid_aodv:

• This one is again similar to above. Just that the movement is not coordinated. Random. More suited for our scenarios.
• After the movement is done then nodes stay at place and do the messaging fireworks etc.
• This just goes on repeating forever.

manetMultiradio:

• Uses mobile hosts with StaticGrid Mobility :-|
• Hosts have multiple radio interface cards. 
• Since there is no mobility,only fireworks happen.

multiradio:

• Could not figure out the diff between manetMultiradio and multiradio

net80211_aodv:

• Contains fixed and mobile hosts. Mobile hosts are not moving but scattered randomly on the playground.

net80211_control:

• Same as could not figure out the difference :-|

[Tanay Upadhyaya]

We all have to prepare some basic stuff about OMNET++, but it won't be possible for everyone to go through the complete routing algorithm for OLSR. The simulation part we should divide amongst ourselves in a way that we almost cover the protocol and different QoS matrices for it in the simulation. So, each can take like 2 of these and vary corresponding parameters to produce their unique study of the protocol. In this, we need to go through only our part of the relevant routing code.

What say?

--
 
 

[Tanay Upadhyaya]

Instead of using the network topology of /examples/adhoc/* in INET, we should use /inetmanet/src/nodes/adhoc/MobileManetRoutingHost.ned. The first one obviously NED file obviously shows a network with many nodes and the second one is just defining the network model, but I'm failing to understand the fireworks thing etc.; it should be possible to make a complete topology from this NED file. Got this course slide by simple Google search which talks about simple simulation using OMNET++. I'll try that tonight to see how close it is to what we need for now. I'm attaching the slide with this mail.

[Siddharth Shah]

So do we meet sir today or tomorrow? 

[Raj Joshi]

We can meet today for anything random we want to ask him. Delivering demo and presentation today seems inhuman !

On 26 September 2012 14:11, Siddharth Shah <siddharth...@gmail.com> wrote:

So do we meet sir today or tomorrow? 

--
 
 

--
Raj Kiran Joshi

4th Year, B.E. (Hons.) Computer Science

+91 978 500 3330

[Siddharth Shah]

Are you in the lab then? We should get together and discuss what we're going to do.

[Raj Joshi]

Yupp since morning :-)

Raj Kiran Joshi

On Sep 26, 2012 2:16 PM, "Siddharth Shah" <siddharth...@gmail.com> wrote:

Are you in the lab then? We should get together and discuss what we're going to do.

--
 
 

[Raj Joshi]

I am running the Grid_Aodv example with just 3 mobile hosts. It is making better sense now. You all may try the same for your respective objectives.

[Tanay Upadhyaya]

We should complete our respective tasks today itself and should meet up tomorrow so that everyone understands the complete demo.

--