I want to design a basic fi-wi network. The idea is to design a fiber network GPON and then add wireless remote nodes (later on implementing WMN). i want to see the feasibility, integration and conversion from wired to wireless network and analyze the network w.r.t various parameters and optimization techniques in wirleless network. i m not sure how much of physical, mac or network layer my work is going to be. i will also be using mathematical modeling for optimization.
Can you suggest which simulation tool should i use omnet++ or ns3 for Fi-Wi so that i can do all this work and also guide me which layer my work will be on mostly?
mnet++ is better choice on physical layer. If you want to implement something on Network Layer then go for NS3,
thanks Asghar for your response. i am yet to learn about my research area everyday and selecting a tool is an important step in it. can you tell me whether omnet does not support all three layers or ns3 does not support the physical and mac layer?
Dear Qadir,they are specialized for this specific layers,However having support for all the layers.
Thanks alot Mr. Asghar. I appreciate your kind response. i myself am more interested in using the omnet, the reason because my main research will focus on decision making and switching between nodes which i have to implement on mac layer and the analysis of the physical characteristics i.e. BER, synchronization, data rate etc. i can perform on physical layer. yet i was confused with network layer implementation.
I would have investigated the available libraries in both in search for the one closest to the protocol stacks you want to use. In the past I've used ns-2 a lot and the large libraries there was its main asset. The current status of ns-3 I'm not aware of, but later Omnet++ w/ INET seems to have a lot onboard. Besides, Omnet++ is very structured when it comes to managing simulations (parameterization/runs/etc).
Both simulators are great alternatives, the election should be based in aspect like, what simulator use the people of my group, and is it he model that I need implemented?
If in your group everyone use ns3 you should use ns3, you will have problems and dudes in your research, to have someone with knowledge about the simulator close is something important. Equally, that the model that you need is implemented in the simulator, it will accelerate your research.
If your model is implemented in both, nobody in your environment work with any of them, the learning curve is better in omnet.
Thanks Arne and Alfonso, i myself am considering omnet++ for one of the reason that few people are already working in it. however, i have yet to explore more about my work.
Since i am a newbie to both the tools so my first goal at this stage is to learn omnet and implement a basic Fi-Wi network and study the behavior of the architecture w.r.t physical characteristics (i.e. data rate, synchronization, BER, delay etc) and at mac layer (how the switching from wired to wireless medium and between nodes will take place) at this stage i kept my focus limited to two layers only and that's why not thinking about routing protocols.
Hi,
I hope you are doing well.
Try to check Opnet (Riverbed), I recommend watching some video for Opnet on YouTube to have a good idea, the work on Opnet is very easy while providing deep visibility into performance.
NS2 and NS3 they are very good and open source as well but the good thing is if you can find some existing code (work was done before - close to your work) that will help you very much.
I strongly recommend trying use each one of them as soon as possible to be clear which one will satisfy your desires of your work.
all the best.
Thanks Mohammed, One of the main reason me or anyone will go for omnet or NS is because its free compared to the price of riverbed modeler (or old opnet modeler.
Regards,
- Badges
- Science topic