Toward High Bit Rate LoRa Transmission via Joint Frequency-Amplitude Modulation

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

1. References

Ensure adding references from recent years. The last reference is from 2023.

2. Abstract

There are some formatting and typographical issues. Please revise the overall content of the titles and subtitles.

Example: In line 4 of the Abstract:

• Original: “a novel PHY-layer scheme that it can transmit data in both amplitude and frequency dimension simultaneously”
• Corrected: “a novel PHY-layer scheme that can transmit data in both amplitude and frequency dimension simultaneously”

3. Formatting and Typographical Issues

Please revise the overall formatting of titles and subtitles.

Example: In section 5.2, “Envaluation Results” should be “Evaluation results”.

4. References and Citations

Please revise references and citations to respect the MDPI requirements.

5. Figures and Illustrations

There are many figures, a total of 20. Are all the figures important?

6. Technical Questions
7. The experimental results presented in the paper focus primarily on short-range scenarios, with most performance evaluations conducted at distances up to 100–150 meters. However, one of LoRa’s key advantages is its ability to support long-range communication over several kilometers. Why were longer-range tests not included in your evaluation? Can you comment on how QR-LoRa would perform beyond 150 meters, especially in rural or wide-area deployments where LoRa is typically used?

 

1. QR-LoRa seems to need special hardware and software, which may not work with existing LoRaWAN devices and gateways. The paper doesn’t explain whether QR-LoRa can fit into current LoRaWAN networks or how it would behave if used alongside standard LoRa devices. Can you clarify whether QR-LoRa is meant to work within the existing LoRa ecosystem, or is it intended to be a completely separate system? How do you see it scaling in real-world deployments where interoperability is important?

 

1. Your paper shows that QR-LoRa can improve data rate and energy efficiency in theory, which is great. But your analysis assumes perfect conditions and doesn’t consider real-world challenges—like how amplifiers might distort the signal, how accurate the ADCs are, or how noise can affect the envelope shape. Since LoRa’s strength comes from using signals that stay the same in amplitude (which helps it work well in noisy or fading environments), doesn’t using amplitude modulation make it more fragile? Do you think QR-LoRa can still work well in real, large-scale IoT networks where devices are cheap, have limited power, and can’t be perfectly calibrated?

 

1. QR-LoRa introduces several additional processing steps—like envelope modulation, denoising, differential analysis, and the MADA algorithm—which likely increase the system’s complexity, power usage, and hardware requirements. However, the main benefit highlighted is a 2× increase in bit rate, which other, simpler techniques like ICS-LoRa or layered CSS also try to achieve. Given this, do you believe the added complexity of QR-LoRa is justified? How does its cost-benefit trade-off compare to these existing, less complex methods?

 

7. English and Grammatical Remarks

• A thorough English proofreading is recommended. The paper has many grammatical errors.

Examples:

In line 4 in the Abstract: “a novel PHY-layer scheme that it can transmit data in both amplitude and frequency dimension simultaneously” → should be “a novel PHY-layer scheme that can transmit data in both amplitude and frequency dimension simultaneously”

• In line 386 in conclusion section: “We achieve the design…” → better: “We present the design…”
• Some paragraphs are too dense; please try to break long paragraphs into shorter ones.

Comments on the Quality of English Language

Please revise the overall formatting of titles and subtitles.

Example: In section 5.2, “Envaluation Results” should be “Evaluation results”.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

1. LoRa technology excels at low-power, long-distance communications, but what are its main bottlenecks? Why is it challenging to increase the data rate of LoRa? Why can combining frequency modulation and amplitude modulation increase the data rate of LoRa? How does this joint modulation approach overcome the limitations of traditional LoRa?
2. What are the core design principles of QR-LoRa? How does it transmit data in both the frequency and amplitude domains?
3. The authors mention that different amplitude coding strategies can be used. What is the impact of different coding strategies on the performance of the system (e.g., data rate, bit error rate)? How to choose the coding strategy that best suits a specific application scenario?
4. Despite the increased data rate, how does QR-LoRa ensure that energy efficiency is not greatly affected? In what cases may QR-LoRa reduce energy efficiency?
5. Why is a cyclic right-shifted ramp signal used for amplitude modulation? How do the specific properties of this signal make it suitable for LoRa systems? How does QR-LoRa's demodulation algorithm handle multipath fading and noise in real channels? How are the proposed de-noising algorithms and the minimum assisted decoding algorithm (MADA) implemented in practical applications?
6. Frequency modulation and amplitude modulation have different noise immunity capabilities. How does QR-LoRa balance these two modulation methods to achieve the best overall performance? How does QR-LoRa adjust its modulation strategy in low signal-to-noise ratio (SNR) and high signal-to-noise ratio (SNR) environments?
7. How is QR-LoRa compatible with the existing LoRa infrastructure? Does this approach require major modifications to existing LoRa devices?
8. In practical applications, what technical challenges may be faced in implementing QR-LoRa? How to solve implementation problems caused by channel characteristics and hardware limitations?

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

Item	Page	Line Number		Comments	Require Response	Severity Level
		Fig 3, 4, 5, 6,	Amplitude vs Time	You need to state the unit of Time ( sec, msec) even if it is normalised	Yes	Moderate
		Fig 8, 9, 10, 11, 12	Amplitude vs Frequency	You need to state the unit of Frequency( Hz, kHz) even if it is normalised	Yes	Moderate
		Fig 15, 16, 18, 19	BR vs. SF measured results BR vs. distance measured results Are these figures for LOS and/or NLOS	You need to state clearly the LOS/NLOS scenario status for each performance plot.	Yes	High
	15	Fig 18		The QRLoRa_amplitude curve fluctuation compared to that of the QRLoRa_frequency indicates that there will be information (SF) misalignment within both the amplitude-domain and frequency-domain, which ultimately reduces the overall BER of the proposed receiver. Hence (if you agree), what you mentioned on lines 77- 80 page 2, “(iii) The anti-noise ability of the amplitude-modulated signal and frequency-modulated signal is different due to the inherent properties of radio propagation, which results in the asymmetry of communication range. We further optimise the noise processing to improve the communication range of amplitude modulation.” è needs to be modified, because you mentioned only the impact on the communication range/link aspect. I suggest that you add something like: the BER (SNR calculation) dependency of the QRLoRa_amplitude is not similar to that of the QRLoRa_frequency, which ultimately reflects that the design needs to address and balance the competing requirements to mitigate such design optimisation misalignment.	Yes	Moderate
	16	391-392	We believe QR-LoRa is a significant step in a line of works that will improve the bit rate gain to 2× compared with the existing LoRa device.	Change to (if you agree): We believe QR-LoRa represents a significant step in a line of work that could improve the bit rate gain by a factor of 2 compared to existing LoRa devices.	Yes	Moderate

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 4 Report

Comments and Suggestions for Authors

The work is interesting and the paper is a good read and brief. I have the following suggestions to improve the quality of the manuscript:

1) Please expand all acronyms at first appearance in abstract then text separately. "LoRa" first appearance in abstract is not expanded. 

2) While your paper pays great deal of effort analyzing noise immunity in amplitude noise, I could not find mention to phase noise (jitter) Please make a comment on phase noise susceptibility based on these 2 references:

[1] A. Hajimiri and T. H. Lee, "A general theory of phase noise in electrical oscillators," in IEEE Journal of Solid-State Circuits, vol. 33, no. 2, pp. 179-194, Feb. 1998, doi: 10.1109/4.658619

(Your noise analysis is very similar to ISF from [2] please give it a look).

3) In Fig 17. NLOS performance is similar to LOS even though there is about 20dB of loss. It seems no multiple path was included in this measurement. I think multiple path is an interesting and more realistic case to consider given the non-zero cross correlation. Can you make a comment on this? Please try to produce a measurement (or a sim if this is too difficult to capture) to study your proposed technique. 

4) Can you comment on the requirement of TX hardware linearity with your proposed technique? Generally the power amplifier (PA) has limited BW, and efficiency, and max power. Does this technique stress the PA? is DPD needed?

 

5) Please consider adding a table where you compare your proposed technique to state of the art works and include numbers and a suitable Figure of Merit (FOM). It would be great to characterize cost. 

6) Recent hardware works use spread spectrum (code domain) for several advantages, Please make a short comment and how feasible can this technique help:

[2] C. Hill, A. Hamza, H. AlShammary and J. F. Buckwalter, "Watt-Level, Direct RF Modulation in CMOS SOI With Pulse-Encoded Transitions for Adjacent Channel Leakage Reduction," in IEEE Transactions on Microwave Theory and Techniques, vol. 67, no. 12, pp. 5315-5328, Dec. 2019, doi: 10.1109/TMTT.2019.2951565

 

7) Can you comment on the pulse shape trade off (e.g. rect vs raised cosine)?

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 5 Report

Comments and Suggestions for Authors

The paper deals with the improvement of classical LoRa technology, consisting in the design and implementation of QR-LoRa, which solves the problem of packet collisions occurring in LoRa. QR-LoRa is a novel PHY layer scheme that can transmit data in both amplitude and frequency dimensions simultaneously.

In a logical sequence, the authors discuss in detail the subproblems and their solutions (e.g. noise suppression), energy consumption and the limiting distance of signal propagation.

They illustrate everything with figures and formally by mathematical equations, and conclude with a series of graphs comparing the important parameters of LoRa and QR-LoRa, where the advantages of the proposed approach are confirmed.

Formally precise is the derivation proving that “QR-LoRa can reduce energy consumption and improve energy efficiency compared with the conventional LoRa.” This could be reformulated into a mathematical theorem.

The authors conclude by emphasising that “QR-LoRa is a significant step in a line of works that will improve the bit rate gain to 2× compared with the existing LoRa device”.

However, a number of other improvements to LoRa technology are mentioned in Introduction, e.g., SSK-LoRa, PSK-LoRa, FBI-LoRa, and so it would be useful to compare QR-LoRa with these approaches.

Less substantial comments:

• On page 3, the same text appears twice: in lines 119-127 and 128-135.
• Page 10, line 243: The BER abbreviation used here is explained only on page 12.
• Page 11, l. 256: "Nyquist Sampling Theorem" – It is frequently referred to as "Nyquist-Shannon Sampling Theorem", "Whittaker-Kotelnikov-Shannon Sampling Theorem", etc.
• Page 12, line 296: "ENVALUATION"? – probably should be "EVALUATION"

 

Typography:

• The differential symbol "d" in Equation (1) should be written in normal style and not in italics (like the functions sin, cos, log, ...).
• The same applies to the exponential function "e" (in Equations (2), (3), (4), (7), (8), (9), (10) and also in the free text).
• The designation BW for “frequency sweep bandwidth” is not very appropriate because it looks like a product of two variables, e.g. $B_W$ would be better.
• Page 5, line 137: Instead of $F^0$, $F_0$ should be.
• Page 7, Equation (7): $s_{F_0,T_0}(t)$ should be to the left of the "=" symbol (similar to Equation (6)).
• Page 9, Algorithm 1: A(i+1), number 1 should be in normal style and not in italics.
• Page 10, Figure 11: A(n) – $A(n)$.
• References to literature should be separated by a space from the preceding word.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The authors provided a satisfactory response to my remarks.

Author Response

Dear Reviewer,
Thank you very much for your thorough evaluation and the constructive response to our manuscript. We highly appreciate the time you invested in reviewing our work and addressing each of our revisions with such care. Your insights have undoubtedly strengthened the manuscript, and we believe the current version is more robust as a result. Please let us know if there are any further details you would like us to clarify—we remain committed to ensuring the work meets the journal’s standards.

Best regards,

Mr. Gupeng Tang

Reviewer 4 Report

Comments and Suggestions for Authors

Thank you for addressing most of my comments. However, two points were not adequately handled:   

1) Phase Noise:  Your response did not meaningfully address the issue. As a reviewer, part of my role is to push you outside your immediate research comfort zone. If you are claiming “strong anti-noise ability” (as stated in the second line of the abstract), then phase noise — particularly its impact on timing recovery and demodulation — must be acknowledged. Please add the reference I suggested in R1 [Hajimiri & Lee, “A General Theory of Phase Noise in Electrical Oscillators”, JSSC, 1998], and include a brief discussion on how noise amplitude fluctuations near LO zero crossings can translate into phase errors and affect QR-LoRa’s performance. Even a qualitative vulnerability analysis would be helpful. 

2) Pulse Shape:  This is a less critical issue, but still worth clarifying. If you have not yet studied the impact of different pulse shapes, at least state what pulse shape was used in your implementation. You can mention that exploring other shaping methods (e.g., raised cosine) is left for future work.     

I look forward to seeing a more complete response in the next revision.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 3

Reviewer 4 Report

Comments and Suggestions for Authors

all my comments are answered now. Thank you.