Could someone tell me the accuracy and credibility of Cooja simulation's results? More published papers utilized Cooja as their experiments and evaluation works since 2012. The easy-to-use toolsets like Cooja/Contiki will be a good alternative for the IoT R&D.
In my view, one may perceive an increased use of Cooja/Contiki for IoT/WoT research. For specific applications such as IoT/WoT, simulation results using Cooja/Contiki are accurate and credible.
On the other hand, NS3 and Omnet can be used for carrying out simulations for networks in general. If proper libraries are available, one may also use them for IoT/WoT. If no such library is available, either one needs to write the library himself/herself for NS3/Omnet or one can simply use Cooja/Contiki.
Totally agree with Ash Abbas's answer, the proper libraries are so important for starting our research more easily. There is no optimal choice currently.
Based on our investigations, current IoT research domain is still at the step of proposing single problem solution, except smart grid and some other few applications.
Actually, this question may be not a so good to clarify the essential subquestion. As we know, Cooja contains emulators (MSPSim and AVRora) so maybe we can extend this question as:
Will one simulator + emulator become a trend of the future evaluation tool-set for LoWPAN and low power and lossy network (defined by IETF ROLL) in IoT ?
Thanks for your appreciations.
One of the reasons behind using either a simulator or an emulator (or a combination of both) is to ensure that the performance of a scheme or a protocol is satisfactory before actually deploying the network, which may require a lot of cost to be incurred. For example, it may happen that one has deployed the network and the performance of the protocol is not satisfactory i.e. the protocol or scheme is required to be modified. In that case the cost involved in deploying the network is wasted. To avoid that one may evaluate the performance of the protocol or scheme using a proper simulator/emulator to make sure that it will perform adequately after its deployment. Good simulator + emulators create environments and scenarios that are relatively closer to the actual deployment scenarios. In my view, one simulator + emulator is sufficient as a tool set for evaluating any scheme or protocol in any network (including LoWPAN and low power and lossy network (defined by IETF ROLL) in IoT). An analytical modelling may further add the technical depth to the contribution.
Dear Yibo Chen
Thank you for your scientific message about Simulation of IoT.. I have M.Sc. student . we need to work on security of "IoT" using hybrid cryptosystem.. The main question for us currently is that : What is suitable for us as ( simulation tool) to be used in such project>
Note: We are working in Babylon University- Iraq. And as I know, there are very small number of academics interesting in IoT in Iraqi Universities... And I hope according to your suggestion, we can go ahead with suitable simulator. Best wishes
@Eng. Sattar B.Sadkhan, you are welcome, sir. In fact, right now, Cooja is more popular than the time when I proposed this question, and you can do any simulation work as you want. Except more hardware emulation supporting, Contiki 3.0 brings much needed support for 802.15.4 link layer encryption with AES128, so maybe the current built-in encryption in Contiki/Cooja will be a good start for you.
- Badges
- Science topic