Measuring the Realtime Capability of Parallel-Discrete-Event-Simulations

Round 1

Reviewer 1 Report

I think the paper represents a nice survey of techniques for Parallel Discrete Event Simulations. I recommend acceptance. Some typos: line 10: "an"-> "a" line 11: remove "and" line 34: "in the area of" line 59: "this metric" line 169: "none"->"non"

Author Response

Dear reviewer,

first, we would like to thank you for your comments. We have conducted extensive spellcheck using the professional Grammarly spell-check tool. We have fixed all the found issues throughout the paper.

 

Reviewer 2 Report

This work presents a review of existing conservative and optimistic synchronisation schemes for PDESs.
It introduces a metric to compare their real-time capabilities to determine whether they can be used for real-time simulation.
An example scenario was simulated in a common, non-parallel DES environment.

Some more detailed comments are given below. I hope that if the authors will take them into account the paper will be improved.
1. The authors have put a lot of efforts on the DES in section 2 and PDES in section 3, but the contribution of the paper is not very clear. The authors can comment on some of the research gap in the field and then introduce the importance or contribution of this work in section 1.
2. There are many journal paper concerning the simulation scenarios designed for VANET from the last two years.
Authors should survey and discuss these state-of-the-art in Section 1.
3. It is hard to link the Real-Time PDES with the Multi-Component System in Section 5. A simple practical example is needed to provide an illustration about the Multi-Component System in section 2.1.2.
4. There are too many references. The authors are suggested to discard some unnecessary and out-of-date references, such as those only showed once in the manuscript.

Author Response

Dear reviewer,

thank you for your constructive feedback.

Please see the attached document for our detailed response.

Author Response File: Author Response.pdf

Reviewer 3 Report

The authors dealt with a topic of interest in the field of real-time simulation. Is made an in-depth analysis of the architectures related to this subject. The content is related strongly with the subject. There is an extensive review about the subject regarding the first four chapters. The theoretical part is sustained by example applications to trough a use case to show the real-time capability. The results are putted in evidence with numerical evaluation. Recommends that this paper be published. As the authors also mentioned, a very important application can be in HIL. For example, in the automotive field where there are very large investments of hardware in order to simulate the whole car as a complete system.

Author Response

Dear reviewer,

we wanted to thank you for your feedback and appreciate your positive comments!

Reviewer 4 Report

Basically, this a review of discrete-event simulator used on simulations with realtime restrictions. I find the paper interesting but I have one big concern that the authors do not explain:

- For the results provided, which is the operating system that runs the simulations? In this case, it is not just the simulator but the operating system. It is not possible to have realtime features if the underlying operating system does not provide them. It does not matter how good is the simulator.

Author Response

Dear reviewer,

we thank you for your constructive feedback.
please see the attached document for our detailed comments.

 

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

The authors have carefully addressed the previous comments of the reviewer and significantly improved the manuscript. 

Author Response

Dear reviewer,

thank you for your positive response.
We are happy to hear that our changes improved our paper.

Best regards

Reviewer 4 Report

After reading the authors' answers I would like to see that they do argue in the text. At least, a short discussion on the real time issues. They have just provided an answer to me but have not included anything in about real time issues in the text. In fact, soft and hard real time are very different things. If authors use soft real time they should say that in the paper. I would like to see things like the operating system used for the test and other features that are significant for the presented results.

Author Response

Dear reviewer,

thank you for your additional comments.

We regret that our previous iteration did not meet your expectations.
Since we agree with you that this discussion will improve our paper, we have added Section 6.2.2 (lines 597 - 610) which addresses the "Influences of the Operating System" with respect to the metric's significance.

Thank you again for your constructive feedback which helps us to improve our paper.

Round 3

Reviewer 4 Report

I still have the same doubts regarding the real time claims. Real time is not a matter of computation power but a matter of guarantees. Indeed, the softness of the real time is mentioned at the end of the paper and just because it is included in a table. I suggest that the title must be changed to include the word soft and also this softness clearly stated in the abstract and the introduction. Also, I expect some discussion in the paper about this issue. It can be argued or not, but soft real time is not real time. The point with soft real time is that with enough resources it is always going to work.

Author Response

Dear reviewer,

we thank you for your sustained encouragement to improve our paper with your constructive feedback.

We have added section 2.1 about the differentiation between soft, firm and hard real time according to literature (please see lines 86-115). Additionally, we restricted the metric to firm and soft real time in the abstract (line 7 and 9). Section 6.2.2 was rewritten (especially lines 633-639) to make clear that the metric adds no further information to systems with hard real-time guarantees. This fact is also stated in the conclusion (lines 677-681).

We hope that this discussion helps to clarify that this metric is used in the context of firm and soft real-time discrete event simulations. No additional information would be added to the simulation system when it is already executed in hard real-time. As you have pointed out, hard real-time systems do not require measuring delays, as they already guarantee that the delays do not exceed certain thresholds.

Again, we want to thank the reviewer for his time to help us to improve the paper.

Best Regards
Christina Obermaier (on behalf of the authors)