Scalable Dew Computing

Round 1

Reviewer 1 Report

The abstract is too long. Please shorten it. 

Author Response

The abstract is too long. Please shorten it. 

Thanks for the comment, now the abstract is shortened.

 

Reviewer 2 Report

Dear Author,

Thanks for submitting this article. It is a very relevant topic and the overall idea is of interest not only to the research community but also to SMI willing to innovate in the IoT domain.

Aspects to improve:

* figures are repetitive and can be condensed into one or two conveying the same message.

* The use case does not properly address/illustrate the scalability argument.

* Research questions are: Is it feasible? Is the resource consumption linear? (these two summarize the three proposed).

* Missing references and discussion for the horizontal offloading methods. The justification of Distributed Horizontal scalability and Centralized Horizontal scalability is poor. They lack detail and must be supported by references to previous studies.

I hope this helps on improving the paper. 

Regards,

Anonymous reviewer.

 

Author Response

Thanks for submitting this article. It is a very relevant topic and the overall idea is of interest not only to the research community but also to SMI willing to innovate in the IoT domain.

Aspects to improve:

Thanks for fruitful comments, please find the response:

* figures are repetitive and can be condensed into one or two conveying the same message.

Fig. 1, 2 and 5 are now being joined into one figure.

* The use case does not properly address/illustrate the scalability argument.

We have included more use-case examples to illustrate the essence of the scalability argument.

* Research questions are: Is it feasible? Is the resource consumption linear? (these two summarize the three proposed).

We have reformulated the first two challenges according to the suggestions, but have not excluded the third, since it covers an aspect that is not covered by the first two.

* Missing references and discussion for the horizontal offloading methods. The justification of Distributed Horizontal scalability and Centralized Horizontal scalability is poor. They lack detail and must be supported by references to previous studies

We have included more references and discussed more details to elaborate the proposed ideas.

Reviewer 3 Report

This paper presents methods to implement the scalability on a dew computing level, which includes two types, the horizontal scalability (distributed and centralized) and the downward vertical scalability. However, only one simple use case is utilized to support these methods. To more the results more convincing, it is better to provide more related use cases.

Author Response

This paper presents methods to implement the scalability on a dew computing level, which includes two types, the horizontal scalability (distributed and centralized) and the downward vertical scalability. However, only one simple use case is utilized to support these methods. To more the results more convincing, it is better to provide more related use cases.

Thanks for the comment, we have included more use cases and cited other authors.