DAG Hierarchical Schedulability Analysis for Avionics Hypervisor in Multicore Processors
Round 1
Reviewer 1 Report
The paper puts forward a method for parallel hierarchical scheduling to be used by VM hypervisors in the context of avionic real-time systems. It is divided into three main sections: Section 4 presents a schedulability analysis (theoretical, based on DAG); Section 5 describes architectural of virtualization platform including hierarchical scheduling; and Section 6 evaluates the proposed approach. At the end, the conclusion is that "experimental results demonstrate that the strategy proposed in the paper have better performance."
Regarding originality and novelty the paper is average because it deals with traditional problems in the area of virtualization technology. The proposed solution is based on DAG task model present in the literature.
Regarding significance of content and interest to the readers, I rank the work high, given the importance of efficiency in the context of real-time application in avionics.
The main "problems" I point has to do with quality of presentation and scientific soundness of the proposal.
First, presentation. Overall, the text is prolix and rather difficult to understand (at least, to me). I've made extensive annotations in the pdf annexed. The main points: (1) Introduction should end with an overview of how the paper is structured, the rationale for the sections; this is important because the text is lengthy and technically challenging. (2) Preliminaries should be rewritten, paying attention to correct use of set-theory conventions and terminology; Equations should be properly numbered; see details in the pdf annexed; (3) Related work is rather superficial; only one paragraph with four citations; should be extended; (4) Schedulability Analysis is the most problematic section; The English is bad, hindering comprehension; Some propositions are named "Theorem" without proper required proof; Some relations are named "Regulation" without proper explanation; The formal development present is excessive and not clearly motivated; What is the point? I think, the authors are supposed to rewrite, re-structure the whole section focusing on the main points/algorithms; (5) Hierarchical Scheduling in hypervisior, on the other hand, is a rather short section 1 + 1/2 page long without proper explanation or motivation; (6) Results is an important section, that make strong statements such as "This results proves the usefulness of the proposed method."; the main problem is the English that deserves a thorough revision; (7) finally, conclusion can be extended to summarize more specifically the main contributions of the work.
Second, scientific soundness. The paper presents both theoretical and practical results. On the theoretical side, the authors state some "Theorems" summarizing properties of the proposed models/approach. Scientifically, to be accepted, theorems require explicit proof. I do not see proofs in the text for the stated theorems. So, I cannot be accepted as scientific sound/valid presentation. Either the "Theorems" should be properly proved or the corresponding properties explained without the name "Theorem". On the practical side, the results are difficult to assess (to me) mainly because of the presentation that is confusing, with many punctuation, grammar and typographical issues. The section must be rewritten.
For more details, see the pdf annexed.
Comments for author File: 
Comments.pdf
Author Response
Dear reviewer，
I am very grateful to you for taking the precious time to put forward your valuable opinions for my paper, and your opinions have been revised in the corresponding places in the paper.
Please see the attachment.
Author Response File: 
Author Response.docx
Reviewer 2 Report
The author presented interesting research results on scheduling in chosen parallel system.
The presented results are important in real time systems. It would be interesting for the reader if the system can guarantee hard real time system functionalities, or the presented method is better than previously known but without guarantee of execution in given time.
The first time used abbreviations should be explained (even the basic one as DAG, CPCM, RTA) in text.
Author Response
Dear reviewer，
I am very grateful to you for taking the precious time to put forward your valuable opinions for my paper, and your opinions have been revised in the corresponding places in the paper.
Thank you very much!
Author Response File: Author Response.docx
Round 2
Reviewer 1 Report
I believe that the points I raised in the first review were duly addressed by the authors in this second version of the article. No more comments.
