Multiuser Equalizer for Hybrid Massive MIMO mmWave CE-OFDM Systems

Round 1

Reviewer 1 Report

This paper presents a hybrid way to design multiuser equalizers. Although the paper considers the hardware aspect, it does not present any hardware-related measures such as the chip area and the power consumption. For the digital part, in particular, it would be possible to count the number of multipliers and adders required. How will the iterative nature affect the total processing time? Besides, inferred from the followings, the manuscript seems to be copied and pasted without proofreading.

1. In Abstract, “archi-tecture”

2. Throughout the paper, the bracket for reference, i.e., [] are omitted, making hard to read.

3. In line 250, the equation is not clearly presented.

Author Response

Author Response File: Author Response.docx

Reviewer 2 Report

The topic dealt by the authors is very interesting, as it addresses a new and very promizing waveform format, the Constant-Envelope OFDM. Such a waveform will have a role in 5G systems based on mm-wave transmission, as claimed by the literature.

The combination of CE-OFDM and massive MIMO is potentially attractive, but also challenging. Authors addresses some important challenges of such integration process. In particular, the use of CE-OFDM in a massive MIMO-assisted uplink that is a concrete novelty. For me, the contribution might be sufficient for publication, but some modifications are anyway needed. They are summarized as follows:

- In this paper, I didn't find any significant insight about the most significant issue of CE-OFDM detection, i.e.: the nonlinear phase demodulation based on arctangent operation. It is known that nonlinear phase detection (present in the receiver architecture of Fig.2) produces noticeable nonlinear effects on the postdetection noise that vary with the modulation index. I don't think that this issue has no impact on the performance of the proposed multiuser equalizer. Authors should clarify this point;

- The analog precoder of section 3.1 is very briefly introduced. Authors should better detail it and insert it in the system framework;

- Why did the author consider the raised cosine pulse shaping? This is not usual in the literature about CE-OFDM. Did the author consider any sidelobe power reduction issue justifying the choice of the RSC? If it is so, please clarify;

- The BER expression of eq. 12 is not well explained and/or justified. Please, check the BER approximations for CE-OFDM presented in the literature (e.g. in Thompson and Proakis paper) and add some clarifications and insights;

- Computational complexity aspects should be clearly detailed in the dealing.

Finally, as minor remark, it would be good to improve the figure readability, in particular for what concerns Fig. 1 and Fig. 2. 

Author Response

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

The authors have well addressed the concerns of the reviewer. Thank you.

Reviewer 2 Report

The author properly answered to my comments. The paper is high-quality and innovative now. It can be accepted for publication.