Want some suggestions

[K B]

I'm new to research area. I have a question that, in any simulator how researchers compares other routing methods with their own proposed routine method. How do they get existing routing method script/ source code ? A very few researchers provide their code in published paper. So, how do new comers compare the existing  routine method ? 

Anyone, please reply !

Thank you.

[zeena usama]

i don't think you need to re-simulate the existing methods or routing protocols

you can do your own routing method and get a result then compare it with the results of the existing method from their research papers or thesis results 

you can also contact the authors if you need more information about their methods  and the results they got..

best regards 

[K B]

Thanks for your valuable reply.

But research papers are come with graph of comparison with existing method not with result of their method. I have tried to contact with author of published papers, but much times get no reply.

Do you know any resource from which I can find source code of any existing method ?

Thank you.

[zeena usama]

Try to search in "github" for the method you are looking for may you find it there 

It have alot of codes and methods for different things 

--
You received this message because you are subscribed to a topic in the Google Groups "OMNeT++ Users" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/omnetpp/NLDwql69Ilg/unsubscribe.
To unsubscribe from this group and all its topics, send an email to omnetpp+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/omnetpp/1332f443-196d-4e4b-8d6a-2c9674bbda43n%40googlegroups.com.

[Charles Pandian]

1. As a simple rule, never trust the results of a paper. Often, scholars always cheat journals, their supervisors and their university with fake/cooked-up results.
2. Frame your research problem only after testing/running/repeating some of the existing algorithms/models (that will be available in the default installation of a simulator - or available as an extension module for the simulator).
   
3. Never trust a algorithm/model that you see on a published paper until you are able to implement it (from the available sources) and evaluate it and getting almost equal results that are published in the original paper.
   

In fact fake publication is common in all the domains of research. But, it is very very high in the CS/ECE/IT community - If you notice previously published work by the "computer science" related domain, you will find a lot of fake publications (publications based on fake results or cooked-up results or hypothetical models).

For example, you may find lot of fake "algorithms" or publications in UWSN, particularly, if they used ns-2/aquasim as a simulation tool. Of course, the original publications of the software aquasim were genuine.  But, by default, all the mac layer protocols and network-layer protocols available in aquasim will not even work on a high-density network (for example 50 or 100 nodes). The reason is the protocol implementations available in aquasim are only at the elementary level (or not at all tested on a  high-density network) - even if you try to move a single node, then all the nodes will just move to 0,0,0 - and the entire simulation will not even work at all. Even the latest ns-2 version of aquasim itself contains a lot of bugs that will lead to wrong results. One of the major bug in aquasim is : if you will try to use a node other than the 0th node as a sink, then the packets will never get delivered to the sink node. 

But if you search papers that are using aquasim, you will see lot of paper that are using algorithms lilke "PSO", "ACO", "GA" and even AI/machine learning-based ones to solve "mobile" UWSN using aquasim -  even the protocols readily available under aquasim will not work as per the original theory (because of some bugs/limitations in ns-2 version of aquasim).  You may understand the genuineness of those papers if you really try to run an existing protocol under aquasim.

Similarly, you can see a lot of fake publications on WSN, VANET and almost all the domains of networking.  

Even you can find fake papers published on repute "top-rated" journals.
No journal review mechanism has a "foolproof" mechanism to detect "fake results" in a paper during its review process.
An intelligent scholar can easily fake things in a paper.

Fake publications can be eradicated if and only if the "scholar/researcher" himself/herself decides to publish a paper only with genuine models and results.

So, fake publications will live until the death of the universe - only you should learn to recognize them and avoid using them in serious scholarly research work.

You may understand about "fake publications" if you are reading the following posts:

‘IEEE Projects’ – A ‘Pitfall’ in front of Real Research and Innovation.
Fake Papers and Damaging Face of Indian Research
The Impacts of a fake publication on a Good Research
Fake Journals – A Humiliation to Scientific Research

Charles Pandian,

--
You received this message because you are subscribed to the Google Groups "OMNeT++ Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to omnetpp+u...@googlegroups.com.

[Rudolf Hornig]

Just jumping in to share out thoughts (I don't want to hijack the thread) :

We (OMNeT++ Development team) see this as a big problem. We would love to see reproducible results in published papers using OMNeT++ . We think that if a paper's result is not repeatable by any reader, then that work by definition has nothing to do with scientific method.

Of course, we cannot force anybody to create such a paper, but we are thinking about ways, how we could make that happen. In the upcoming versions of OMNeT++ we will provide recommendations and maybe some example skeletons, how to set up such a project. We hope that in the future, this would be the NORM and every reviewer will automatically check for this. But as a first step, we have to make this easy for the authors and then to the reviewers... This will be a loooong game.

Rudolf

[Alfonso Ariza Quintana]

My two cents about this,

I usually receive a couple of works to review for clustering routing-based protocol proposing increasing the lifetime of the nodes. Most of them use Matlab to simulate the model and basically, all of them use a non-realistic transceiver consumption model.

They all assume that the transceiver only consumes when it transmits or receives, and the consumption in the transmission is a lot bigger than in reception, and finally the consumption in transmission is based on the free space model, but the electronic doesn't work as they propose.

First, the transceiver consumes if it is connected, it is more, we have measured some commercial transceivers cc2420 and similar, and we have measured the peak precisely in the idle state because in this state the transceiver has the amplifiers at the maximum to increase the sensitivity, but, when you read the papers, in this state, they assume that the consumption in this state is zero. Some authors have a bit more idea about MA protocols and propose to use a TDMA protocol, but the problem about how to assign the time slots or how to synchronize the nodes they ignore. Others propose a complex algorithm that knows the position of the nodes, be when I asked about how they know the position they answered that using GPS, but the problem is that a GPS transceiver consumes more than the 802.15.4 transceiver, after this, I commented that all the consumption results that they presented are false and impossible to achieve.

The LEACH protocol was published in 2000, I have read many clustering LEACH-based protocols, but I have never seen a real network that uses this type of protocol and I haven't found a measure of the lifetime in a real network using clustering-based protocols

---

De: omn...@googlegroups.com <omn...@googlegroups.com> en nombre de Rudolf Hornig <rudolf...@gmail.com>
Enviado: sábado, 18 de junio de 2022 19:33
Para: OMNeT++ Users <omn...@googlegroups.com>
Asunto: Re: [Omnetpp-l] Re: Want some suggestions

 

To view this discussion on the web visit https://groups.google.com/d/msgid/omnetpp/2e50ef44-42bb-4c9c-8101-3651933a2520n%40googlegroups.com.