An Interference-Aware Resource-Allocation Scheme for Non-Cooperative Multi-Cell Environment

Round 1

Reviewer 1 Report

Authors propose An Interference Aware Resource Allocation Scheme for Non-cooperative Multi-cell Environment. 

According to the reviewer opinion, declared "Multi-cell Environment" is considered only for special case of one and two interferers, as it is depicted in Figure 1. If it is not the case, authors shoul detail simulation and validation setup layout more precisely. Declared "Multi-cell Environment" should be substantiated.

Strong side of investigation is prototyping setup with USRP boards. However, described layout also details only  special case of one and two interferers. 

Weak sides of investigation can be  improved, if authors consider the following suggestions:

1. Algorithm 1 DQN-based joint sub-channel allocation and transmission rate selection for brevity should include references to formulas instead of verbal description.

2. During simulation, transmission period is set to be T = 100 time slots. For example, in LTE it is equal to 5 frames with duration of 50 ms. How does it account UE mobility and the ability of proposed solution to promptly respond to Rayleight fading during coherence time?

3. References to figures should precede figures itself. Plotted notations of Reward convergence and Episode (x10) should be somehow explained.

4. Simulation conditions should be described in details. Is it link or system level simulation? For the case of interference account it should be system level simulation, thus, how many base stations and UEs are accounted? What is the terrestrial layout, transmiter powers, relative distances, antenna radiation patterns, inter-site distances and so on. If it is just a special case, depicted in  the Figure 1, it is still a lot of parameters, to be accounted, when evaluating interference.

5. Convergence of proposed solution depends on CSI and feedback channel which in OFDMA is also subject to interference in multipayh fading. Authors are suggested to indicate the level of abstraction, at which simulation takes place.  How simulation results can be verified, is it available somewhere, which simulation tool was utilized?

6. How does prototype scenario reproduces Rayleight fading, used in simulation? May be it is LOS conditions, which is also just a special case?

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Inter-cell interference has a negative impact on cell-edge users and system performance, and reduces network capacity. In this manuscript, Interference cancellation has been studied based on historical channel variations and interference characteristics. The authors proposed an interference-aware prediction-based resource allocation strategy for non-cooperative multi-cell transmission. The authors suggested collecting and storing the historical noisy channel states to predict the upcoming channel condition and determined the efficient resource allocation strategy accordingly. This study is well explained, the manuscript is interesting to read. However, there is only one shortcoming needed to resolve before publication.

-The authors proposed a model for interference cancellation and defined the model suitably. However, adding an explanation, in the introduction, about the reason that caused the interference in the system is required.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

In this paper, an interference-aware prediction-based resource allocation strategy to deal with multi-cell interference has been proposed. In addition, the historical noisy channel state and the acknowledgement feedback are fully utilized. The reviewer feels that the overall paper is structured reasonably, references are cited correctly, experiment results are presented, and conclusions are consistent with the evidence and arguments. However, the reviewer feels several points must be addressed in the revised manuscript.

 My suggestions and recommendation are as follows.

 General comments:

 

1. In order to improve readability author can write "In [ref]" instead of [ref]. For example, "In [13] ...sentence..." for instead of [13] ...sentence.... Also, please maintain consistency in all places.

2. Please increase the font size of the text in Figure 4, Figure 9 and maintain consistency in all figures if possible.

3. In order to improve the readability and clarity, authors may consider replacing Fig. 6 (b) with screenshot of the laptop and maintain consistency in other figures where possible, for example Figure 8.

Technical comments:

1. The reviewer feels that the novelty and contribution of the presented work were not highlighted correctly. To highlight novelty and contribution, authors can summarize key results in the table and compare them with the literature.

 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Authors provided detailed answers to all the six suggestions in their response, and for reviewer it is enough. However, in the latest version of manuscript additions, highlighted in blue, seems not enough.

For example, multi-cell scenario substantiation and appropriate restrictions becomes convincing from the figure in the response, but not from the original figure in the revised manuscript; authors could add full figure from the responce and comments to it in the final version.

According to the reviewer opinionin, authors should include detailed answers in the responce to the latest version of manuscript in appropriate sections. 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf