kumoh national institute of technology
Networked Systems Lab.

Review Comment

NSL > Works@NSL> About Review> Review Comment
Godwin Brown Tunze, Huynh-The, Jae-Min Lee, and Dong-Seong, " Multi-shuffled Convolutional Blocks for Low-complex Modulation Recognition, " 2020 ICTC, Oct. 21 - 23, Jeju, Korea (A)
By : Godwin
Date : 2020-09-14
Views : 46

--------------- Reviews ------------------------------------

======= Review 1 =======

> *** Relevance: How well does the content fit the conference scope? Is this paper handling an important theme in this area?
Good (4)

> *** Completeness: Does this paper describe the problem clearly? Are the results of this paper reproducible via experiments (implementations, proofs)? How well is the result analysis done with the previous works? How clear is the paper's conclusion for the problem tackled?
Excellent (5)

> *** Originality: Does this paper include any novel approaches or new applications that have never been tried?
Excellent (5)

> *** Presentation: Are the title, abstract, and keywords appropriate? How proper is the organization and description method of this paper?
Good (4)

> *** Comments to authors: Please provide detailed comments to the authors.

This paper proposed a convolutional neural network architecture for the automatic modulation recognition (AMR) in resource-constrained devices. The architecture consists of a series of shuffled blocks with residual connections. A channel shuffling module is deployed to improve the communication among filter groups. A residual connection is deployed from an input of each shuffled block to its corresponding output through an element-wise additional layer to learn the spatiotemporal features from in-phase and quadrature signals repetitively to enhance the recognition accuracy. The performance of the proposed scheme is evaluated through experiments with the RadioML2018.01A dataset. The results show that it can achieve high recognition accuracy and trainable parameter utilization efficiency over the state-of-the-art approaches.

> *** Recommendation: Please provide your overall recommendation on the acceptance of the paper. (Final acceptance decisions will also consider literal responses to the questions below.)
Strong Accept (5)

======= Review 2 =======

> *** Relevance: How well does the content fit the conference scope? Is this paper handling an important theme in this area?
Excellent (5)

> *** Completeness: Does this paper describe the problem clearly? Are the results of this paper reproducible via experiments (implementations, proofs)? How well is the result analysis done with the previous works? How clear is the paper's conclusion for the problem tackled?
Excellent (5)

> *** Originality: Does this paper include any novel approaches or new applications that have never been tried?
Good (4)

> *** Presentation: Are the title, abstract, and keywords appropriate? How proper is the organization and description method of this paper?
Good (4)

> *** Comments to authors: Please provide detailed comments to the authors.

A CNN architecture for automatically recognizing modulation is suggested in this paper. The reason why the specific AI tools was mentioned well and the simulation results substantiated the efficiency of the proposed method.

> *** Recommendation: Please provide your overall recommendation on the acceptance of the paper. (Final acceptance decisions will also consider literal responses to the questions below.)
Strong Accept (5)

======= Review 3 =======

> *** Relevance: How well does the content fit the conference scope? Is this paper handling an important theme in this area?
Good (4)

> *** Completeness: Does this paper describe the problem clearly? Are the results of this paper reproducible via experiments (implementations, proofs)? How well is the result analysis done with the previous works? How clear is the paper's conclusion for the problem tackled?
Good (4)

> *** Originality: Does this paper include any novel approaches or new applications that have never been tried?
Good (4)

> *** Presentation: Are the title, abstract, and keywords appropriate? How proper is the organization and description method of this paper?
Good (4)

> *** Comments to authors: Please provide detailed comments to the authors.

The paper deals with the interesting topic for the conference. Look forward to seeing the presentation during the conference.

> *** Recommendation: Please provide your overall recommendation on the acceptance of the paper. (Final acceptance decisions will also consider literal responses to the questions below.)
Accept (4)