eWoT: A Semantic Interoperability Approach for Heterogeneous IoT Ecosystems Based on the Web of Things

Round 1

Reviewer 1 Report

This paper addresses an important problem in IoT field. Due to numerous standards and implementation, IoT devices would not compatible with another device. RDF-based thing description would help to relax the difficulty, and its distributed architecture makes it efficient to use.

The proposed approach makes sense; however, it could be further improved, considering the follows:

-. There are numerous interface designs in IoT systems. Proposed eWoT is essentially an additional indirection layer that manages the IoT devices. Good thing after having the eWoT is obvious. However, it assumes that the devices support eWoT; Because most existing IoT devices do not support eWoT, it is not going to be compatible with them. Can it be possible to incorporate heterogeneous IoT devices that do not support eWoT?

-. eWoT provides distributed access to devices' sensor data, in order to provide low latency. The effectiveness can be analyzed in various ways; the current paper presents only a few example cases, based on simulation. I could not find flaws in the simulation; yet, it is hard to derive a general conclusion such that the proposed protocol is efficient. Even with the given application, I can hardly draw out the conclusion that it can effectively support heterogeneous IoT devices. 

Author Response

Dear reviewer

 

This paper addresses an important problem in IoT field. Due to numerous standards and implementation, IoT devices would not compatible with another device. RDF-based thing description would help to relax the difficulty, and its distributed architecture makes it efficient to use.

 

Thank you very much for the kind comments.

 

The proposed approach makes sense; however, it could be further improved, considering the follows:

-. There are numerous interface designs in IoT systems. Proposed eWoT is essentially an additional indirection layer that manages the IoT devices. Good thing after having the eWoT is obvious. However, it assumes that the devices support eWoT; Because most existing IoT devices do not support eWoT, it is not going to be compatible with them. Can it be possible to incorporate heterogeneous IoT devices that do not support eWoT?

 

This point has been clarified in page 7, paragraph "Task 1: Registering an IoT devices", making it explicit how a new device becomes compatible with eWoT.

 

-. eWoT provides distributed access to devices' sensor data, in order to provide low latency. The effectiveness can be analyzed in various ways; the current paper presents only a few example cases, based on simulation. I could not find flaws in the simulation; yet, it is hard to derive a general conclusion such that the proposed protocol is efficient. Even with the given application, I can hardly draw out the conclusion that it can effectively support heterogeneous IoT devices.

 

We have included a new experiment that uses a third-party data API from an European project, i.e., Stars4ALL. This API publishes the data of different photometers (different from the sensors used in the experiment 2 which were smart energy meters for houses and appliances like washing machines, air conditioning, etc.) sparse across the world. In this context we executed eWoT demostrating that SPARQL queries can be resolved using the data provided by the Stars4ALL API.

In addition, in this context we implemented a centralised approach in which a third-party software periodically copied the values of the different sensors published by means of the Stars4ALL API into a graphDB repository. Then, the same queries used to evaluate eWoT were evaluated, and the results compared.

Finally, for the camera ready version of the paper a proof reading will be carried out to solve grammar mistakes.

 

Kind regards,

 

Andrea C., María P.V. and Raúl G.C.

Reviewer 2 Report

This paper proposes a novel approach that builds semantically interoperable ecosystems of IoT devices. Overall this paper presents a solid and complete work. However, this paper didn't present the state of the art of relevant topics. Some important references are missing. A more comprehensive literature review of relevant topics is desired. Some example references are: Sabina Jeschke, Christian Brecher, Houbing Song, and Danda Rawat, Industrial Internet of Things: Cybermanufacturing Systems.  ISBN: 978-3-319-42558-0, Cham, Switzerland: Springer, 2017, pp. 1-715; J. Gao, X. Yang, Y. Jiang, H. Song, K. R. Choo and J. Sun, "Semantic Learning Based Cross-Platform Binary Vulnerability Search For IoT Devices," in IEEE Transactions on Industrial Informatics. doi: 10.1109/TII.2019.2947432

Author Response

Dear reviewer,

 

This paper proposes a novel approach that builds semantically interoperable ecosystems of IoT devices. Overall this paper presents a solid and complete work.

Thank you for your kind and encouraging comments.

 

However, this paper didn't present the state of the art of relevant topics. Some important references are missing. A more comprehensive literature review of relevant topics is desired. Some example references are: Sabina Jeschke, Christian Brecher, Houbing Song, and Danda Rawat, Industrial Internet of Things: Cybermanufacturing Systems.  ISBN: 978-3-319-42558-0, Cham, Switzerland: Springer, 2017, pp. 1-715; J. Gao, X. Yang, Y. Jiang, H. Song, K. R. Choo and J. Sun, "Semantic Learning Based Cross-Platform Binary Vulnerability Search For IoT Devices," in IEEE Transactions on Industrial Informatics. doi: 10.1109/TII.2019.2947432

We have included the references mentioned in order to complete the coverage of the paper, references [5] and [41].

 

Finally, for the camera ready version of the paper a proof reading will be carried out to solve grammar mistakes.

 

Kind regards,

 

Andrea C., María P.V. and Raúl G.C.

 

Reviewer 3 Report

The proposed approach should be compared with other papers that offer similar solutions

Author Response

Dear reviewer,

 

We have included a new experiment that uses a third-party data API from an European project, i.e., Stars4ALL. This API publishes the data of different photometers (different from the sensors used in the experiment 2 which were smart energy meters for houses and appliances like washing machines, air conditioning, etc.) sparse across the world. In this context we executed eWoT demostrating that SPARQL queries can be resolved using the data provided by the Stars4ALL API.

In addition, in this context we implemented a centralised approach in which a third-party software periodically copied the values of the different sensors published by means of the Stars4ALL API into a graphDB repository. Then, the same queries used to evaluate eWoT were evaluated, and the results compared.

The comparison with other proposals from the sate of the art is not a trivial task  due to the heterogeneity of these proposals and the lack of implementations availability. For instance, the Centralised approaches require further development of code. Due to this reason we implemented a basic centralised solution that follows the description provided by the literature for this kind of proposal (referenced in the paper). Nevertheless, this problem of comparing proposals has led us to think that a framework to fair compare this proposals is required.

Finally, for the camera ready version of the paper a proof reading will be carried out to solve grammar mistakes.

 

Kind regards,

 

Andrea C., María P.V. and Raúl G.C.

Round 2

Reviewer 3 Report

Although the proposal includes three type of experiments, this research should be compared with other existing proposals in the state of the art. In page number 8, it is not clear why eWoT scales linearly. The scalability process should be better explained

Author Response

Please see the attachment.

Author Response File: Author Response.pdf