The Role of Artificial Intelligence in Next-Generation Handover Decision Techniques for UAVs over 6G Networks

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Summary:

This manuscript surveys handover (HO) decision and optimization techniques for cellular-connected UAVs, with a particular emphasis on AI/ML methods and their potential role in future 6G ecosystems. The topic is timely and relevant, and the paper contains useful background material and a taxonomy of AI/ML approaches.

 

Comments and Required Revisions:

1、

The paper claims a unified taxonomy and comparative framework, yet the comparative synthesis is mostly qualitative. Please add at least one comprehensive comparison table that summarizes representative works by: (i) scenario (non-UAV, UAV-BS, UAV-UE), (ii) learning paradigm (SL/USL/RL/DRL/FL/etc.), (iii) input features (RSRP/RSRQ/SINR/load/mobility), (iv) optimization targets (HOPP, RLF, throughput, latency, energy), (v) evaluation environment (simulator/testbed/field), and (vi) reported gains over baselines.

2、

The section on 3GPP measurement events and HO control parameters is helpful, but the manuscript should better connect standardized mechanisms to the AI/ML methods surveyed. A structured mapping from controllable parameters to learning-based decision variables (state/action/reward design) would strengthen the technical depth.

3、

Sections 4 and 7 are both organized around “challenges,” which makes them feel repetitive and blurs the chapter-level responsibilities. Conceptually, Section 4 reads more like engineering/phenomenon-level HO challenges and mechanisms, while Section 7 should emphasize open research gaps and barriers to reproducibility/deployment. However, the manuscript does not provide a clear boundary between the two, and Section 4 lacks a closing paragraph that “funnels” the engineering issues into research gaps. In addition, Section 7 would benefit from a bridging artifact that explicitly maps each phenomenon/problem in Section 4 to the corresponding research gap (e.g., data scarcity, generalization, real-time constraints, reproducibility obstacles) and to potential AI/ML directions. I recommend: (i) adding a concise transition/summary at the end of Section 4 that reframes the discussed HO issues into concrete research needs; (ii) opening Section 7 with a short paragraph clarifying its distinct scope relative to Section 4; and (iii) including a Section 4 issues/Section 7 gaps/barriers /candidate AI/ML approaches mapping table/figure. Consider renaming the two section titles to be more distinguishable.

 

Author Response

Dear Editors,

Please find attached the revised manuscript with highlighted changes, along with our responses to reviewers’ comments in the same file.

Thank you for your kind consideration.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

Data & methodology: 

• It is recommended (not mandatory) to create a table summarizing references and results, which will facilitate readers in tracking the existing studies.
• IMPORTNANT: The manuscript includes a section on AI/ML approaches (5.2.1. to 5.2.3), but the general terms for AI, ML, DL, SL, USL, and RL are described inconsistently. In my view, this section could be shortened as it primarily covers general theories about AI. Additionally, the authors discuss the Fuzzy Logic approach in section 5.2.4 without referencing it earlier or clarifying that it is not considered part of AI/ML. In my opinion the authors should create better classification of Handover Decision Techniques for UAVs over 6G Networks.
• Is it appropriate to include this number of level three chapters when each contains only a single paragraph?

References: 

• Paragraph 6.2 does not clarify whether publicly available datasets are available for Handover Decision Techniques for UAVs over 6G Networks.

Clarity and context:

• TYPOS: line 35, 465-467 (blanks).
• TYPOS: lines 112, 115, 119, 132, 465, 466, 467, 737, 781, 897, 912, 941, 985, 986, 1028, 1029… (written w/o space between references)
• The formatting of Table 1 is poor.
• It is advisable to revise subtitles 3.1.2., 3.2.2. and 3.3.2., currently titled "ML–Based Methods," as "ML" encompasses a broader range of algorithms than those discussed in this section. The authors specifically refer only to Deep Learning Supervised Learning (DL) and Reinforcement Learning (DRL) methods. Furthermore, since the text subsequently employs the AI/ML or similar terminology, it is advisable to maintain consistency by using uniform terms throughout.
• It is recommended to provide a step-by-step explanation of the scenario presented in Figure 6.
• ISR is not defined.

 

Author Response

Dear Editors,

Please find attached the revised manuscript with highlighted changes, along with our responses to reviewers’ comments in the same file.

Thank you for your kind consideration.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

There are following observations about this manuscript:

1. The organization of the manuscript must be revisited such as there should be clear illustrations about research questions(RQ), research aims and objectives. In the current structure there is a gap how this research has progressed without having explicit roadmap about starting phases of the research. Figure 1 is only representing the structure of the manuscript. There should be detailed flow chart to highlight the flow of research conducted.
2. Table 1 need to be expanded with more parameters.
3. Figure 3 is very generic need to be revised.
4. Again Table 2 is very short and need to be expanded.
5. Section 3, Existing Studies  must have one table for comparison of published work in each sub headings.
6. Section 4 must have scenario diagram.
7. Figure 6 need to explained properly.
8. Section 7 must be accompanied with proposed solutions.
9. Conclusion is too long it should precisely highlight the key facts.
10. Perhaps above the conclusion there should be critical discussion section.

Author Response

Dear Editors,

Please find attached the revised manuscript with highlighted changes, along with our responses to reviewers’ comments in the same file.

Thank you for your kind consideration.

Author Response File: Author Response.pdf

Round 2

Reviewer 3 Report

Comments and Suggestions for Authors

Although authors by mistake submit  the revised manuscript in the author response section however after carefully checking the revised manuscript file, it is clearly reflected that authors have addressed the queries, therefore manuscript may be accepted for the publication in the present form.