Joint Spectrum and Power Allocation with User Association for 5G/Wi-Fi Coexisting Millimeter Wave Networks

Round 1

Reviewer 1 Report

Authors have proposed a Joint Spectrum and Power Allocation with User Association for

5G/WiFi Coexisting Millimeter Wave Networks at 60GHz. Paper is written well. I have the following comments.

1.       CSMA/CA is not defined.

2.       Why Nr is higher than Nt?

3.       What is the impact of the number of users on the throughput after 20 users? Kindly highlight this point in the revised manuscript.

4.       Kindly include the convergence of algorithm-2.

5.       Word MINLP is not defined.

6.       Qos should be QoS.

7.       Define AoA/AoD

8.       Kindly add more details of the simulation results.

9.       Kindly improve the reference section. Authors may include some recent reference like

Non-orthogonal multiple access: An enabler for massive connectivity  and

Precoded spatial modulation-aided cooperative NOMA 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

 The authors consider the problem of maximizing the uplink system data rate in a hot-spot gathering scenario with 5G and WiFi coexisting millimeter wave networks.

1) While WiFi and cellular will coexist with each other, it is necessary for the authors to outline in this research the unique benefits for each when serving millimeter wave specific applications.

2) Building on 1), the authors need to emphasize the limitations of unlicensed spectrum and Wifi. They also need to acknowledge other solutions that can enable solving the problem of congested networks.

3) It is not clear why the authors consider the uplink system sum data rate, rather than an end-to-end approach that considers the downlink and the uplink.

4) The approach deployed, i.e., use a distance-based user association approach can be problematic in mmWave given that blockages are known to be an important challenge of mmWave links. The authors should consider looking at the SINR instead.

5) The comparison to baselines adopted seems a little problematic. The authors first formulated an optimization problem and then divided it to three subproblems. However, if one looks at the baselines considered, it is apparent that the other two (or one) subproblem is not solved. Thus, it is intuitive that MDRA will lead to superior results.

6) Building on 5) the authors need to find another technique that can solve the formulated problem and then compare their results.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

- The discussion of the related works needs to be elaborated.

- Simulation results are noisy (curves are not smooth enough). How was the simulation done? 

- Simulations results are not enough to show an original contribution by this work. What about other performance parameters?

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

N/A

Reviewer 3 Report

Most of the remarks have been addressed. I have no other comments.