Full-Duplex MAC Protocol Based on CSMA/CA for Switching Transmission Mode

Round 1

Reviewer 1 Report

One of the important reasons that control network performance is to determine the transmission mode for transmitted and received packets in the network environment. The FD is one of the important modes in the network that has brought the attention of researchers and academics to develop it. However, this transmission mode still has some problems. In this paper, the authors proposed a hybrid algorithm (FD/HD transmission) to improve transmission mode performance depending on MCS level. They claimed that their proposed determines a better transmission mode, which in turn improves the network's throughput capacity and reduces latency. In addition, they designed an analytical model to evaluate the hybrid transmission mode based on FD pair probability. With a network having 40 nodes, they pointed out that their proposed algorithm (hybrid transmission) reduces the latency of delay by 10% and 13% compared to existing research. The topic is significant because of its link with network performance, throughput, and packet exchange. Authors must address all of the following comments/questions carefully.

• We suggest that the authors should remove keywords such as “Full-duplex” and " MAC protocol" because these keywords are already found in the title of the article. It is better that they replace them with other terms to increase the reach of the paper.
• Did the authors calculate the delay amount for choosing a transmission mode (HD/FD/Hybrid)? How?
• If there is a network that contains a number of APs and some of these APs work with HD transmission mode, is this mean that the transmission and reception of packets will affect the performance of the entire network? Existing research has indicated that a network that uses a single transport mode (HD or FD) for APs has higher throughput than a network that uses different modes for APs.
  • Lee, J., & Quek, T. Q. (2015). Hybrid full-/half-duplex system analysis in heterogeneous wireless networks. IEEE transactions on wireless communications, 14(5), 2883-2895.
• What is the probability that the Tgain value is less than or equal 0 and what affects the network throughput (Algorithm1, Fig2(C))?
• Authors should add a comparison table that includes parameters, limitations and strengths between the proposed and the existing research to clarify the superiority of the proposed research.
• Authors pointed out that “FD transmission rarely occurs when SIC = 80 dB” (page10-lines173-174). What is a scientific and fair interpretation to explain that the proposed does not achieve superiority over the existing research, both in terms of throughput and delay (when using SIC (dB) =80 (Figure 8 (a) and (b)).
• Conclusion section: This section does not clearly describe the conclusions extrapolated from the study. The authors should improve this section. Also, we suggest adding future directions in this section.
• All figures are drawn with high precision/resolution. However, Figures 1 and 2 are shown prior to their use in-text. The authors do not indicate and explain the use of Fig2(a) in-text. Similarly, Tables 1 and 2 are shown prior to their use in-text.
• References list: The number of references is insufficient for this study. This research requires adding recent research (Especially in the Introduction and system model sections).
• Proofreading: This article has some minor typos and grammar. Some of the words require "a", "an" and "the" in different places in this paper (such as candidate (page1-line24), data (page2-line51), lowest (page8-line141), … etc.). Authors should add “,” after “Finally” (page2-line72), replace “detail” with “detailed” (page5-line105), many acronyms that were used before authors define it (such as B5G, UL, DL, MCS, WLAN, … etc.).

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

I found this work interesting, but there are tree aspects that, in my opinion, should be included in the text. Two are relative to the protocol and one to the simulation.

My first question if the backoff procedure, after the transmission of a packet, unless the node is working in Txop mode (HCF), the node must program a Backoff procedure and affect the delay and performance. In the analysis the backoff procedure is not considered, it should be interesting to include some information about it, if affect the protocol or, it can be removed, or when is necessary to use.

The second question is relative to the ACK. The standard is quite strict with the time that a node can wait the reception of an ACK, but with the modifications I can see that it is possible to delay the ACK or after the transmission of a frame the next received frame could be the other frame different to the ACK, by default, the standard programs a retransmission, information how the authors solve this problem, how to inform to the node that the ACK can be delayed or that the next frame could be not the ACK should be included. It is possible to increase the time that a node can wait an ACK, but in this case the performance will be degraded if the probability of lost a frame is relatively high.

My last question is relative to the simulation model, it is not clear is the authors are using a pure analytical model or a discrete event simulation. If the authors have used a simulator, if the simulator is the 802.11 model presented in the WLAN toolbox, if the simulation is at the system level, if they use abstraction physical layer model ….

If they are using a simulation at the system level, there are much information that should be included, traffic, landscape size, mobility model, number of repetitions with different seeds ………

Also, if the results are obtained using simulations, the confidence interval should be included.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

The authors proposed a hybrid transmission switching algorithm in case of asymmetric FD pairing.

The paper needs to be enhanced by reflecting the followings mainly for system model;

  - Detailed procedure including RTS/CTS, measurement of signal and interference, SIR calculation, SIR reporting with multiple nodes involved.

  - N nodes are considered. But the paper did not consider the matching. It should be clearly explained how the UL/DL pairing is performed. 

  - Since the AP cannot know the UL/DL FD matching in advance, SIR calculation and reporting need to be done for all the other nodes except the transmit node.

  - In case of hybrid transmission, UL/DL transmission do not overlap then, no interference. Then the SIR is no longer required. Then, SNR is used? This kind of procedure is stated? The estimation of SNR in addition to SIR needs to be explained.

  - The hybrid transmission is the same as the HD except for DL data in the middle of transmission. In Fig. 2 (a), the DL data transmission should be included for better comparison. And this comparison/difference between (a) and (c) needs be addressed in the context as well. 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 3 Report

The paper has been revised reflecting all the comments properly.