Adaptive Modulation Scheme for Soft-Switching Hybrid FSO/RF Links Based on Machine Learning
Round 1
Reviewer 1 Report
Comments and Suggestions for Authors
This paper investigated the modulation adaption problem in hybrid FSO/RF links based on AI algorithm machine learning model. By deducing the link budget under snow and fog weather condition, the authors gave the link margin and SNR threshold for different order PPM/PSK/QAM schemes. And some experiment was given with a local weather data set by AI training and testing process. The simulation result in this paper can give some good reference to implement soft switching hybrid FSO/RF communication links. But there are also some weak spot and grammar mistakes listed as follows, which should be modified carefully.
1. In eq(6), eq(8) and eq(9), the units for weather attenuation formulas were not given, describe it clearly to make the later calculation understandable.
2. The abbreviation "LM" in Tables 2 and 3 is not mentioned before in the paper, present it clearly.
3. In lines 180 and 181, how to chose the target BER? what's the selection principle, more details show be given here.
4. In lines 164, 165 and 225, check the "cumulative distribution function" number usage, align with the standard quantifier.
5. In lines 191-192, the author should specify the amount of data used for two links when dealing with rainy and foggy weather, and supplement the FSO rainy data.
6. In Figure 6, why the modulation schemes decided in test results were fewer than those in the training results? Give the reasons and analyze them in detail.
Comments on the Quality of English Language
The quality of English language is well.
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for Authors
L42: target and the -> target, and the
Table 1: add space between the numbers and units in Value column
L194: Table 3 -> Table 4
Also in Table 4, it confusing to have columns with same name twice, please clarify this
General: It's somewhat awkward to call "training result" as typically you don't do the inference at training stage.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 3 Report
Comments and Suggestions for Authors
Dear Authors,
Here are a couple of suggestions to improve your manuscript:
1. Some spelling errors. Worth a thorough review.
2. Labellings in all Figures are hard to read, i.e. they are small.
a. Fig 1 better high-resolution labellings.
b. Same thing in Fig 2.
c. Figs 3-7 are all graphs needing larger axes labels.
Best regards.
Comments on the Quality of English Language
Spelling errors.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
