Cost-Aware Bandits for Efficient Channel Selection in Hybrid Band Networks
Round 1
Reviewer 1 Report
To improve the quality of this manuscript, my suggestions are listed as follows.
(1) The quality of Abstract should be improved. The purpose part is missing.
(2) The literature review, main contributions and paper architecture should be separated in three paragraphs in Section Introduction.
(3) The literature review is weak. Authors should make a completed survey and add some new references.
(4) For the academic paper, we usually use the three-line table.
(5) The discussion and conclusions can be separated into two Sections.
(6) The references are not enough for an excellent literature review.
Author Response
We would like to deeply thank the respected reviewer for appreciating our humble work and found it as very much significance for future communication systems.
Author action: We tried our best to cope with all comments/concerns given by the respected reviewer that we do believe they positively enhance the presentation, clarity, and novelty of the revised paper.
Reviewer#1, Concern # 1: The quality of Abstract should be improved. The purpose part is missing..
Author response: We would like to deeply thanks the respected reviewer for this comment. We totally modified the abstract to reflect the purpose part. The modified abstract in the revised manuscript as given in the author action.
Author action: To cope with this valuable comment, abstract is modified in the revised manuscript as follows:
Revised Manuscript
“
”
Reviewer#1, Concern # 2: The literature review, main contributions and paper architecture should be separated in three paragraphs in Section Introduction..
Author response: We would like to deeply thank the respected reviewer for this valuable comment. We modified the introduction section and included related work and paper contributions as a subsections inside it.
Author action: To cope with this valuable comment, we modified the introduction section in the revised manuscript as follows:
Revised Manuscript
“
”
Reviewer#1, Concern # 3: The literature review is weak. Authors should make a completed survey and add some new references.
Author response: We would like to deeply thank the respected reviewer for this deep insight comment. The literature review are updated to up to date related work in the revised manuscript
Author action: To cope with this valuable comment and make the reviewers concern clear for the readers, we have added the following sentences in the revised manuscript:
Revised Manuscript
“
”
References
“
”
Reviewer#1, Concern # 4: For the academic paper, we usually use the three-line table.
Author response: We would like to deeply thank the respected reviewer for this deep insight comment. The table is updated according to the reviewers vision.
Author action: To cope with this valuable comment and make the reviewers concern clear for the readers, we have added the following sentences in the revised manuscript:
Revised Manuscript
“
”
Reviewer#1, Concern # 5: The discussion and conclusions can be separated into two Sections.
Author response: We would like to deeply thank the respected reviewer for this deep insight comment. The discussions are added to the results part.
Reviewer#1, Concern # 6: The references are not enough for an excellent literature review.
Author response: We would like to deeply thank the respected reviewer for this deep insight comment. The references are updated to cover the related work with updating the literature review as previously discussed in your third concern.
Author Response File: Author Response.pdf
Reviewer 2 Report
The article is very well written and structured. The results are quite interesting. However, it seems to me that more state of the art articles should be added, and therefore more references (there are only 11).
Author Response
Reviewer#2: The article is very well written and structured. The results are quite interesting. However, it seems to me that more state of the art articles should be added, and therefore more references (there are only 11).:
Author response: We would like to deeply thank the respected reviewer for the time and efforts she/he spent in reviewing our humble work especially this reviewer gives us wonderful deep insight comments/concerns.
Author action: We tried our best to cope with all comments/concerns given by the respected reviewer that we do believe they positively enhance the presentation, clarity, and novelty of the revised manuscript.
Author Response File: Author Response.pdf
Reviewer 3 Report
Authors in this research form the problem of optimal band/channel selection in hybrid radio frequency and visible light communication (RF/VLC) networks as a multi-objective optimization problem. By solving this optimization problem, the throughput of a hybrid network can be increased while the energy consumption can be reduced. They propose two algorithms, CSUCB-HBS and CSTS-HBS, to solve this optimization problem efficiently.
The manuscript is generally well structured, but can be further improved:
- Justifying the necessity of introducing both CSUCB-HBS and CSTS-HBS
Considering that the key idea in this research is expressed in Equations (8a)–(8f), it may be necessary to provide only one algorithm to solve this optimization problem efficiently. Another algorithm seems redundant. If the authors still want to keep both, they need to justify the advantages of both and how they complement each other.
- Doing real experiments and/or more numerical simulations
Currently, there is just one set of experiments in this research. Doing numerical simulations with another set of parameters can make the proposed method more convincing. Additionally, if possible, the authors may embed the algorithms into a hybrid network and acquire some real data. Showing data from real experiments will substantially increase the reliability of this whole research.
Author Response
Reviewer#3: Authors in this research form the problem of optimal band/channel selection in hybrid radio frequency and visible light communication (RF/VLC) networks as a multi-objective optimization problem. By solving this optimization problem, the throughput of a hybrid network can be increased while the energy consumption can be reduced. They propose two algorithms, CSUCB-HBS and CSTS-HBS, to solve this optimization problem efficiently.
The manuscript is generally well structured, but can be further improved:
Author response: We would like to deeply thank the respected reviewer for the time and efforts she/he spent in reviewing our humble work especially this reviewer gives us wonderful deep insight comments/concerns.
Reviewer#3, Concern # 1: Justifying the necessity of introducing both CSUCB-HBS and CSTS-HBS
Considering that the key idea in this research is expressed in Equations (8a)–(8f), it may be necessary to provide only one algorithm to solve this optimization problem efficiently. Another algorithm seems redundant. If the authors still want to keep both, they need to justify the advantages of both and how they complement each other.
Author response: We would like to deeply thank the respected reviewer for this valuable and constructive comment. Both UCB and TS are stochastic MAB techniques. Hence, the two algorithms are considered to apply the cost subsidy on both algorithms and compare between them too. As results stated CSTS-HBS is better due to the policy of TS that uses Bayesian policy.
Reviewer#3, Concern #2: Doing real experiments and/or more numerical simulations
Currently, there is just one set of experiments in this research. Doing numerical simulations with another set of parameters can make the proposed method more convincing. Additionally, if possible, the authors may embed the algorithms into a hybrid network and acquire some real data. Showing data from real experiments will substantially increase the reliability of this whole research.
Author response: We would like to deeply thank the respected reviewer for this valuable and constructive comment. We will consider that as a future work.
Author action: To cope with this valuable comment, we updated the conclusion of the revised manuscript as follows:
Revised Manuscript
Future work includes multi Tx-Rx scenario extension and further real data investigations.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
I am satisfied with authors’ revisions. But they should pay attention to the following issues before the formal publication.
(1) The title of Section 6 should be changed for Conclusion and outlook.
(2) The font size in Figs. 3-5 is very big.
Author Response
Reviewer#1: I am satisfied with authors’ revisions. But they should pay attention to the following issues before the formal publication.
(1) The title of Section 6 should be changed for Conclusion and outlook.
(2) The font size in Figs. 3-5 is very big.
Author response: We would like to deeply thank the respected reviewer for the time and efforts she/he spent in reviewing our humble work especially this reviewer gives us wonderful deep insight comments/concerns. Both two issues are handled in the revised manuscript as we changed Section 6 name to Conclusions and outlook and also, we reformatted figures 3-5 with suitable font size
Author Response File: Author Response.pdf
Reviewer 3 Report
None of my concerns has been addressed.
Author Response
Reviewer#3: None of my concerns has been addressed.
Author response: We would like to deeply thank the respected reviewer for the time and efforts she/he spent in reviewing our humble work especially this reviewer gives us wonderful deep insight comments/concerns. In this round, we greatly focus on your previous comments and tried our best top reflect it on the revised manuscript.
Reviewer#3, Concern # 1: Justifying the necessity of introducing both CSUCB-HBS and CSTS-HBS
Considering that the key idea in this research is expressed in Equations (8a)–(8f), it may be necessary to provide only one algorithm to solve this optimization problem efficiently. Another algorithm seems redundant. If the authors still want to keep both, they need to justify the advantages of both and how they complement each other.
Author response: We would like to deeply thank the respected reviewer for this valuable and constructive comment. The main idea of this paper is to apply the cost subsidy idea to solve the HBS issue. Hence, we applied it on both UCB and TS, which are two famous stochastic MAB techniques for general comparison. As Algorithm 1 describes both techniques in only one algorithm. This is due to the main difference between UCB, which applies upper bound policy and TS that implements Bayesian strategy. CSTS-HBS have the best performance due to correct prior reward estimation which is Gaussian due to additive white Gaussian noise.
Author action: To cope with this valuable comment, we updated the conclusion of the revised manuscript as follows:
Revised Manuscript
“Envisioned CS-HBS methods”
In this section, we discuss our planned CSUCB-HBS and CSTS-HBS schemes. Both methods modify the naive UCB and TS stochastic bandit schemes, respectively to be cost subsidy, where the Tx can tolerate small loss from the largest payoff (subsidy i.e., the amount of payoff the Tx may forgo to refine costs). Since UCB and TS are the most common stochastic MAB schemes, we straightforward the cost subsidy concept on both of them to check their performance regards HBS issue.
“Results”
Thus, CSTS-HBS arises as the best economic approach for the HBS difficulty due to not only the TS's Bayesian learning policy of TS but also its ability to save energy consumption. The second-best performance is the CSUCB-HBS due to applying both upper bound and cost subsidy policies.
Hence, the promising practical results enhance the affordability of the proposed CSTS-HBS then CSUCB-HBS schemes in HBS in B5G/6G systems.
At t approaches 400, the proposed CSTS-HBS scheme converges to 99.5\% of the optimal throughput due to Bayesian and cost subsidy strategies. Furthermore, the CSUCB-HBS attains 97.2\% due to the upper bound and cost subsidy conceptualizations.
Reviewer#3, Concern #2: Doing real experiments and/or more numerical simulations
Currently, there is just one set of experiments in this research. Doing numerical simulations with another set of parameters can make the proposed method more convincing. Additionally, if possible, the authors may embed the algorithms into a hybrid network and acquire some real data. Showing data from real experiments will substantially increase the reliability of this whole research.
Author response: We would like to deeply thank the respected reviewer for this valuable and constructive comment. In the revised version we added the energy efficiency performance of the proposed HBS approaches versus distinct blocking scenarios (no blockage, small blockage, large blockage). Here the energy efficiency is defined as the average throughput over energy expenditure per selected band in bit/sec/joule. Regards embedding the algorithms into a hybrid network we will consider this step as future work due to time limitations.
Author action: To cope with this valuable comment, we updated the conclusion of the revised manuscript as follows:
Finally, Fig.6 exhibits the viability of our envisioned HBS schemes in terms of energy efficiency (defined as the average payoff (throughput) over energy expenditure per chosen band in bit/sec/joule [33,34]) in Gbps/mJ over distinct Tx-Rx distances. Here, Figs. 6(a), (b), and (c) preview the energy efficiency performance for no-blocking, small blocking, and large blocking layouts, respectively. For all approaches, the energy efficiency is inversely related to the separation distance because of the path loss effect. For all plotted blocking layouts, the CSTS-HBS method outperforms other approaches due to its better performance and appropriate channel selection policy reflecting lowest energy consumption. Meanwhile, TS-HBS,CSUCB-HBS, and UCB-HBS displayed encouraging energy efficiency performances over all separation distances, respectively. However, due to the traditional choice strategies, the random and conventional approaches show much worse performance due to randomization and attempting whole available bands, respectively. Therefore, these promising experimental results ensure that our suggested CSUCB/TS-HBS approaches, especially CSTS-HBS, can be raised as the most feasible HBS method in multi-band wireless communication networks.
Author Response File: Author Response.pdf
Round 3
Reviewer 3 Report
Thank you. My concerns have been addressed.