S.O.V.O.R.A.: A Distributed Wireless Operating System

Round 1

Reviewer 1 Report

Good idea, methodology and realisation. In the paper too many abbreviations.Graphic attachments are difficult to read.
Improve the English language.

Author Response

Thank you for your revision, we improve the paper, the sections: related work, results, language and explain with more details the prototype and the testbed.

We reduce the abbreviations , thank you.

Author Response File: Author Response.pdf

Reviewer 2 Report

General comments

This article (which is of a good scientific interest) is quite difficult to understand due two major flaws :

1) this article is ill-structured and should be better organized. Some key notion are used before being properly introduced, the same notions are sometimes separated by a third element

2) the high usage of acronym (even when they are not necessary) is a bottleneck to the reading.

The authors do not mention what is the objective of the article in the abstract but clearly mention it in introduction : "The main contribution of this paper is the design and implementation of a distributed wireless operating system, dedicated to the dynamic management of computing resource in MANET devices [3]. The goal is finding a solution to dynamically (re-)organize computation tasks on available network resources and node status. The operating system uses two elements to create and deploy virtual resources: abstraction and the  primitives, which are managed by the Local Agent (LA) that resides as an instance on each physical node and the Orchestrator (OR) that work on one or more nodes to monitor and deploy policies on the distributed operating system." Yet when the article comes to experimental setup, no performance setup/framework has bee defined to demonstrate the objective of the articl so the results are quite useless since the authors failed to set clearly what they are aiming at demonstrating.

Last : are the authors dealing with OS or more Middleware? IN the case of OS, the section 2 shoudl also give a state of the art of Operating Systems for dsitrbuted WSN or IoT Fog Computing?

In details :

Structure of article

Section 2 must be segmented in several subpart to clearly structure the state of the art. As presented here, the section 2 is the enumeration of work related to the article but give the reader the feeling that he has himself to do the correspondence, the link with article.
Line 116 is introduced correctly Orchestrator BUT this notion is already used line 94?
The CPS are introduced in this section but it hard to see the relation with ad-hoc network. Futhermore CPS are mentioned, MCPS also but a paragraph dealing with other things(electric vehicle MonteCarlo Chain) separate them (l129-131)?
Last but not least, the article talks about distributed systems but deals in majority with fog computing. A significant part of real wireless distributed system is not dealt here and should be integrated. The reviewer can recommend the reading of the following paper to improve this dimension
[1] E. Di Pascale, I. Macaluso, A. Nag, M. Kelly, et L. Doyle, « The Network As a Computer: A Framework for Distributed Computing Over IoT Mesh Networks », IEEE Internet Things J., vol. 5, no 3, p. 2107‑2119, juin 2018, doi: 10.1109/JIOT.2018.2823978.
[2] S. He, H.-S. Shin, S. Xu, et A. Tsourdos, « Distributed estimation over a low-cost sensor network: A Review of state-of-the-art », Information Fusion, vol. 54, p. 21‑43, févr. 2020, doi: 10.1016/j.inffus.2019.06.026.
[3] O. Zedadra, A. Guerrieri, N. Jouandeau, G. Spezzano, H. Seridi, et G. Fortino, « Swarm intelligence-based algorithms within IoT-based systems: A review », Journal of Parallel and Distributed Computing, vol. 122, p. 173‑187, déc. 2018, doi: 10.1016/j.jpdc.2018.08.007.
[4] D. Datla et al., « Wireless distributed computing: a survey of research challenges », IEEE Communications Magazine, vol. 50, no 1, p. 144‑152, janv. 2012, doi: 10.1109/MCOM.2012.6122545.

The beginning of the third question should be numbered and clearly defined in its objective. We need to wait for the section 3.1 and 3.2 to understand what is not so clearly explained in this introductive paragraph and to understand what is BATMAN or ALFRED. A rewriting of this part is strongly suggested. The reviewer think that section 3.1 and 3.2 and 3.3 should come first to facilitate the understanding of this section.
line 183 the figure 3 is to far (2 pages away) to clearly understand the demonstration.
line 194, the OR is defined as being on a node so where is the distributed approach if the OR is on a centralized paradigm?
Line 256, Linux is mentioned as OS but was never introduced before? Why  the authors are using Linux?
Section 3.4 should be put before and perhaps even in section 2 : MANET is used since the beginning of the article and only defined page 8/16? Section 3 is clearly not the appropriate place to define MANET.

Section 4: line 348 to 358 are clearly the exact repetition of the previous paragraph? The authors should perform a closer reading of their submission to alleviate such a raw mistake.

Scientific discussion

Experiments are not clearly presented : the authors should present the experimental testbed including the protocols for current/power measurement (precision, materials used, …). Furhtermore, nowhere is defined the key performance parameters of their approach so it is quite unclear for the reader what is at stake here : a performance validation (in such a case comparison is lacking) or a feasibility and good working of this approach (in this case the performance set should be clearly asserted in the beginning of the section). As a consequence, discussion of the results are seen as only description without clear assessment or demonstration. In particular in the conclusion the authors say that “this kind of network allows the mitigation of routing problems, node discovery” but it is not demonstrated in the experiment results (where the task repartition is).

 

Author Response

Thank you for your revision, we improve the paper, the sections: related work, results, language and explain with more details the prototype and the testbed.

We reduce the abbreviations , thank you.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

General comments :

The quality of the paper has significantly improved : authors have integrated most recommandations (except one about scientific discussion) of the previous review and implemented satisfying complementary elements to answer most question of the previous review.

Yet, two main parts need to be addressed :

• first the comments of reviewer on scientific discussion have not been adressed and/or not replied in the cover letter. In particular, the part 4 should be modified to adress this part (cf. detailed comments later) and section 5 amended to reflect this.
• second, figures are often referenced far from their physical position in the paper (or not/illed referenced) making it difficult to read sometimes

Scientific discussion has no answer in the cover letter, neither wireless distributed system with suggested references. The authors should specify how the address (or not) the questions here?

Detailed review

Figures are generally ill placed in the paper and some figure/table are not referenced in the document. Figure 1 is page 3 and referenced only two pages later, figure 2 is referenced on top of the page7 while the figure is on the top of the page 6, table 3 and 4 are not referenced in the paper, figure 14 is not referenced in the paper. Figure 18 should be converted in english.

A close reading of the paper should be performed. For example typo line 226 is still present after the first reviewing process, spanish word are used in figure, and so on?

Section 2 is now clearer but some minor revision are needed.
Line 100 (and then line 103), it is not clear what is regular MANET and what is not regular MANET.
Fog computing is more an IoT paradigm than a MANET paradigm, it should be adressed more specifically on the MANET dimension.
CPS section is still unclear as the reason why it is developped here. Why CPS are presented here specifically? They are presented here and never used after in the paper. They do not contribute significally to the state of the art of the paper so why keeping it?
The authors do not perform a review of OS for MANET distributed systems. They should use reference work such as "O. Hahm, E. Baccelli, H. Petersen, et N. Tsiftes, « Operating Systems for Low-End Devices in the Internet of Things: A Survey », IEEE Internet Things J., vol. 3, no 5, p. 720‑734, oct. 2016, doi: 10.1109/JIOT.2015.2505901." to ground their argumentation. Section 2.2.4 is supposed to deal with OS but they deal with frameworks (line 209) and middleware (line 210). If this section aims at dealing with framework for ioT architecture, the title of this section should reflecty that but the reviewer still believe that since SOVORA is an OS for MANET, a state of the art of OS for WSN/MANET should be developped.

Section 3 has been structured but Figure 5 showing the global structure and then modeling of SOVORA should be kept in the beginning to make easier the understanding of the now well comprehensive subsections.

We still need to wait for ligne 356 to understand what BATMAN stand for whereas it is used many times before.

The section 4 should be structured to clearly show what is demonstrated hear : a feasibility with some key performance metrics.

• Unfortunately at the end of the section 4 there is no summary of what has been demonstrated (and it isn't done neither in the conclusion part). This aspect was already notified in the previous review and has not been adressed
• In the section 4 we have some results about power and performance and sometimes we have some extracts showing the sequential process of SOVORA (fig7, fig 8 and fig 14). These log are put in the middle of node consumption making this part unclear (line 420 and then after some results line 444). Perhaps author should structure this part with subpart and separate what is log and process description, from what is testbed specification and scenario hypothesis from what is results. With perhaps a discussion part for the conclusion of this "Experiment and results" section.
• Figure 14 is never referenced is the paper?

Section 5 as conclusion should be rewritten : in particular the reviewer is still not convinced of the mitigation of routing problem as stated in the conclusion and already asked in previous review?