Exploitability of Maritime Fleet-Based 5G Network Extension

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The paper provides a inter-area research and study for the future technology in Maritime application. 
Strength of the paper lies for the new ways of extending 5G relays using the data from the AIS. It uses machine learning and focuses on SAR Logistics, 
Passenger Services, and defense.
It uses a very large AIS records of 1.5B AIS records to analyze ship movement and thereby extracting coverage demand.
The paper also discusses on economic insight, geospatial modeling, policy and regulatory awareness.
However, the reviewer wants throw some light after going through the manuscripts:
1. There is lack of any formal mathematical modeling on how the results have been obtained. Provide model and methodology how the results were obtained from raw AIS message. Provide a sample AIS message and what are the information that has been extracted or extrapolated.
2. There is one mathematical equation and that is not properly labeled.
3. Replace all figure with better resolution.
4. Redundant information must be removed.
5. Paper has been overly described with conceptual information, and multiple parts can either be joined or can be removed, rather more information regarding mathematical and algorithmic procedures can be presented. 
6. Presentation style such in page no. 22, where 5.1 starts with points from no.8, that too in  separate section. This attributes to poor presentation. Similar observation can be seen in 5.2,   5.3. somewhere it 5G, somewhere 5g. Please look into such corrections. 
7. The term 5G is not directly related to the work in the paper. It also creates a ambiguity, can be replaced with "mobile coverage". 

Comments on the Quality of English Language

Please look into the typos and presentation style. Reduce the Irrelevant  content.

Author Response

Comments 1:
1. There is lack of any formal mathematical modeling on how the results have been obtained. Provide model and methodology how the results were obtained from raw AIS message. Provide a sample AIS message and what are the information that has been extracted or extrapolated.

Response 1:
The data used was from preprocessed datasets, the text has been amended to explain that along with the description of data processing done within this study.

 

Comments 2:

2. There is one mathematical equation and that is not properly labeled.

Response 2:

We recreated the formulas and labels.

 

Comments 3:
3. Replace all figure with better resolution.

Response 3:

The figures  might have been quality due to conversions between web and desktop word processors. The images have been replaced with the originals.

 

Comments 4:
4. Redundant information must be removed.

Response 4:

Text has been edited

 

Comments 5:
5. Paper has been overly described with conceptual information, and multiple parts can either be joined or can be removed, rather more information regarding mathematical and algorithmic procedures can be presented. 

Response 5:

Text has been edited

 

Comments 6:
6. Presentation style such in page no. 22, where 5.1 starts with points from no.8, that too in  separate section. This attributes to poor presentation. Similar observation can be seen in 5.2,   5.3. somewhere it 5G, somewhere 5g. Please look into such corrections. 

Response 6:

The text has been amended accordingly.

 

Comments 7:

7. The term 5G is not directly related to the work in the paper. It also creates a ambiguity, can be replaced with "mobile coverage". 

Response 7:

While not explicitly mentioned in the text, other capabilities of 5G such as edge computing differentiate it enough from mobile coverage provided by 2G-4G.

Reviewer 2 Report

Comments and Suggestions for Authors

The manuscript clearly describes the innovation of the low-cost multi-hop model in the naval 5G networks context. However, the strengths can be enhanced in abstract and discussion/conclusion sections. In the abstract, some numerical analysis is missed. The description of the context is well done, in fact the technological gaps between satellites and terrestrial networks is introduced with recent references.

The novelty of the end-to-end approach (hardware/software integration) is not explicitly compared with previous works (e.g. studies on IAB in 3GPP Release 16) and in particular the description of how the model overcomes the bottlenecks of traditional networks is limited and a little bit unclear.   The pros of the proposed method are the massive use of AIS data (1.5 billion points), hexagonal models for spatial analysis, and a clear description of scenarios (Danish Straits vs. Aegean). However, the manuscript can benefit by a workflow or diagrams illustrating the interaction between naval relays, satellites and public/private networks. The formula in Page 16 must be inserted in mathematical format (I think it's a problem with Latex, you should use \begin{equation} \end{equation}). Some parameters are introduced but not correctly justified. Why 1 km² for hexagonal cells? Lack of sensitivity to different resolutions. What is the used optimizer for economic models? Stochastic gradients? The figures are clear and explanatory, but for figures 3 and 4 the legend must be extended, you have to consider figures (6-7-8-10-11-12) with higher resolution for better visualization and it can be helpful for figures 5-9 place the RGB images of the case studies side by side to understand which areas are considered. On the other hand, table V does not define the metrics used (e.g. "95% SAR Coverage" should be correlated to km² or number of incidents/year).

In the comparison analyses a baseline is missed. It can be beneficial a comparison with LTE-Maritime networks or hybrid solutions (e.g. OneWeb + 5G). Other metrics can be helpful. For instance, can you consider Energy efficiency of naval relays (W/Mbps) or environmental impact (CO₂ saved)?

There is a lack of discussion, in fact the limitations are not considered: Performance with bad weather or network congestion (e.g. QoS drops above 50% load). Some concrete proposals for the future developments can be interesting to introduce. (e.g. integration with fuel cell drones or ETSI standardization).

This manuscript can be accepted for its innovative and promising findings, but more technical and comparative rigor are required. Specifically: addition of flowcharts and technical details (e.g. latency per hop); Inclusion of comparisons with 6G/satellite networks and additional metrics (energy efficiency); expansion of the discussion on operational limits and standardization roadmap.

 

Author Response

Comments 1:

The manuscript clearly describes the innovation of the low-cost multi-hop model in the naval 5G networks context. However, the strengths can be enhanced in abstract and discussion/conclusion sections. In the abstract, some numerical analysis is missed. The description of the context is well done, in fact the technological gaps between satellites and terrestrial networks is introduced with recent references.

The novelty of the end-to-end approach (hardware/software integration) is not explicitly compared with previous works (e.g. studies on IAB in 3GPP Release 16) and in particular the description of how the model overcomes the bottlenecks of traditional networks is limited and a little bit unclear.   The pros of the proposed method are the massive use of AIS data (1.5 billion points), hexagonal models for spatial analysis, and a clear description of scenarios (Danish Straits vs. Aegean).

However, the manuscript can benefit by a workflow or diagrams illustrating the interaction between naval relays, satellites and public/private networks. The formula in Page 16 must be inserted in mathematical format (I think it's a problem with Latex, you should use \begin{equation} \end{equation}). Some parameters are introduced but not correctly justified.

Response 1:

The equation might have been lost its formatting due to conversions between web and desktop word processors. We recreated the formula.

Comments 2:

Why 1 km² for hexagonal cells?

Response 2:

The 1 km is edge size and can be considered the resolution of final raster. Thus, it represents the sub-cell on connectivity within which the connectivity remains similar. The text is amended to explain that.

Comments 3:

Lack of sensitivity to different resolutions. What is the used optimizer for economic models? Stochastic gradients? The figures are clear and explanatory, but for figures 3 and 4 the legend must be extended, you have to consider figures (6-7-8-10-11-12) with higher resolution for better visualization and it can be helpful for figures 5-9 place the RGB images of the case studies side by side to understand which areas are considered. On the other hand, table V does not define the metrics used (e.g. "95% SAR Coverage" should be correlated to km² or number of incidents/year).

Response 3:

The table is now amended.

Comments 4:

In the comparison analyses a baseline is missed. It can be beneficial a comparison with LTE-Maritime networks or hybrid solutions (e.g. OneWeb + 5G). Other metrics can be helpful. For instance, can you consider Energy efficiency of naval relays (W/Mbps) or environmental impact (CO₂ saved)? There is a lack of discussion, in fact the limitations are not considered: Performance with bad weather or network congestion (e.g. QoS drops above 50% load). Some concrete proposals for the future developments can be interesting to introduce. (e.g. integration with fuel cell drones or ETSI standardization).

Response 4:

Other metrics are part of future research based on same data. The limitations of scope mean that not everything can be included in this paper. However, we do acknowledge that this is a valid concern and amend the conclusions accordingly.

Comments 5:

This manuscript can be accepted for its innovative and promising findings, but more technical and comparative rigor are required. Specifically: addition of flowcharts and technical details (e.g. latency per hop); Inclusion of comparisons with 6G/satellite networks and additional metrics (energy efficiency); expansion of the discussion on operational limits and standardization roadmap.

Response 5:

The more technical side was omitted due to potential overlap with another publication.

Reviewer 3 Report

Comments and Suggestions for Authors

This paper presents an interesting optimization study of 5G multi-hopping implementation applied in two areas of high maritime traffic, along with a concise survey of maritime networking solutions and some performance tests of multihop 5G network implementation scenaria. Optimization strategies and results are presented clearly and convincingly. I would like to recommend publication, with some minor comments as follows.

• References 1-2 are cited in Section 2 at various points with respect to the survey of current maritime connectivity solutions, but their relevance is not obvious, given that they are focused on High Altitude Platform Stations (HAPS). Perhaps they would be more appropriately cited elsewhere.
• Lines 130-145, referring quite generally to maritime sector economics, are not particularly relevant to the subject of the present work; in my opinion, this text could be ommitted at no cost to the overall quality of the paper
• The authors generally give the explicit definition of acronyms at first occurrence, but there seem to be some slips; e.g., I cannot locate the definition for IAB (integrated access and backhaul), apparently occurring for 1st time in line 318, though derivative mIAB is defined in lines 414-415. (Perhaps being pedantic, I notice the same for 3GPP). Also, a reference for the UERANSIM software (e.g. link to github or wiki) would be welcome.
• lines 309-310: "Thus presents a holistic overview ... in marine technologies" --> obscure phrase, subject missing (who?)
• There is a broken link in line 322
• line 432: "In configuration B, it was implemented" --> somewhat obscure meaning
• point list in lines 889-892: the numbering seems to be incorrect

Author Response

Comments 1:

This paper presents an interesting optimization study of 5G multi-hopping implementation applied in two areas of high maritime traffic, along with a concise survey of maritime networking solutions and some performance tests of multihop 5G network implementation scenario. Optimization strategies and results are presented clearly and convincingly. I would like to recommend publication, with some minor comments as follows.

References 1-2 are cited in Section 2 at various points with respect to the survey of current maritime connectivity solutions, but their relevance is not obvious, given that they are focused on High Altitude Platform Stations (HAPS). Perhaps they would be more appropriately cited elsewhere.

Response 1:

The solutions are not endemic to maritime connectivity; thus it was reasonable to reference meta-analysis already performed by other researchers.

 

Comments 2:

Lines 130-145, referring quite generally to maritime sector economics, are not particularly relevant to the subject of the present work; in my opinion, this text could be omitted at no cost to the overall quality of the paper

Response 2:

We deem the economic overview necessary for background. As such, omitting the text does not provide any improvement.

 

Comments 3:

The authors generally give the explicit definition of acronyms at first occurrence, but there seem to be some slips; e.g., I cannot locate the definition for IAB (integrated access and backhaul), apparently occurring for 1st time in line 318, though derivative mIAB is defined in lines 414-415. (Perhaps being pedantic, I notice the same for 3GPP). Also, a reference for the UERANSIM software (e.g. link to github or wiki) would be welcome.

Response 3:

UEARANSIM reference added. The acronym explanations were lost due to restructuring the article, we added them back.

 

Comments 4:

lines 309-310: "Thus presents a holistic overview ... in marine technologies" --> obscure phrase, subject missing (who?)

Response 4:

Edited, the original sentence was malformed during previous restructuring.

 

Comments 5:

There is a broken link in line 322

Response 5:

There was another in line 344, both are fixed now.

 

Comments 6:

line 432: "In configuration B, it was implemented" --> somewhat obscure meaning

Response 6:

Edited for better readability.

 

Comments 7:

point list in lines 889-892: the numbering seems to be incorrect

Response 7:

Renumbered to start from 1.

Reviewer 4 Report

Comments and Suggestions for Authors

This manuscript employs the Baltic Sea and the Mediterranean Sea as case studies to systematically analyze the availability, economic feasibility, and impact of fleet-based 5G network expansion in maritime environments. It investigates the role of multi-hop 5G connectivity through rigorous cost-benefit analyses and business model validation. Furthermore, it proposes an economic assessment framework for evaluating the feasibility of deploying 5G networks across different maritime routes and recommends policies to promote the development of cross-border 5G maritime networks. The selected topics demonstrate significant application value. However, the following issues require further attention:

1. Section 2.1.3 mentions SDN dynamically managing network links but lacks concrete examples of implementation or performance metrics. It is suggested that additional technical details be provided to strengthen credibility.
2. In the 5G-multi-hop technical framework section, while it is proposed that ships can serve as mobile relay stations to achieve multi-hop 5G coverage, specific protocols or algorithms (e.g., routing strategies, load balancing mechanisms) for multi-hop communication are not elaborated upon. Additional explanations are recommended to enhance clarity and technical depth.
3. The methodology section does not address how to handle spatial and temporal discontinuities in AIS data (e.g., data sparsity when the ship is stationary or at low speeds), nor does it consider the influence of seasonality or extreme weather on navigation patterns. Data preprocessing methods and a thorough analysis of limitations should be included.
4. The results indicate a more efficient base station deployment in the Aegean Sea; however, it remains unclear whether the distribution of islands contributes to a reduction in the coverage area of a single station. Further analyses are required to evaluate the impact of geographic factors on base station deployment.
5. The conclusion highlights that lessons learned from 5G network extensions can be applied to 6G development. However, the adaptability of 6G technical parameters to maritime networks is not adequately addressed throughout the text. It is recommended to revise the conclusion to clearly distinguish between established 5G findings and speculative projections for 6G.
6. The economic feasibility analysis states that the installation cost of the vessel ranges from 70,000 to 100,000 euros. However, it lacks details regarding cost components (e.g., hardware, installation, maintenance) and whether long-term operation and maintenance costs (e.g., software updates, equipment replacement) have been accounted for. Additional clarification is therefore recommended.
7. The numbering of the charts is inconsistent, with Figure 3 appearing twice. This issue should be revised to ensure proper continuity and accuracy in figure labeling.

Author Response

Comments 1:

1. Section 2.1.3 mentions SDN dynamically managing network links but lacks concrete examples of implementation or performance metrics. It is suggested that additional technical details be provided to strengthen credibility.

Response 1:

Section was edited; additional references were added.

 

Comments 2:

1. In the 5G-multi-hop technical framework section, while it is proposed that ships can serve as mobile relay stations to achieve multi-hop 5G coverage, specific protocols or algorithms (e.g., routing strategies, load balancing mechanisms) for multi-hop communication are not elaborated upon. Additional explanations are recommended to enhance clarity and technical depth.

Response 2:
This part was omitted due to internal distribution of research results between working groups and their respective Publications. We have now added additional information along with necessary reference.

 

Comments 3:

1. The methodology section does not address how to handle spatial and temporal discontinuities in AIS data (e.g., data sparsity when the ship is stationary or at low speeds), nor does it consider the influence of seasonality or extreme weather on navigation patterns. Data preprocessing methods and a thorough analysis of limitations should be included.

Response 3:

We added explanations along with references. Since the data covers extensive time periods, the seasonal patterns can be considered averaged out at this scale. However, in future research this is a very important aspect, and we will go in depth on that.

 

Comments 4:

1. The results indicate a more efficient base station deployment in the Aegean Sea; however, it remains unclear whether the distribution of islands contributes to a reduction in the coverage area of a single station. Further analyses are required to evaluate the impact of geographic factors on base station deployment.

Response 4:

Yes, further analyses are needed and will be done in ongoing studies. We added respective comments into the conclusions.

 

Comments 5:

1. The conclusion highlights that lessons learned from 5G network extensions can be applied to 6G development. However, the adaptability of 6G technical parameters to maritime networks is not adequately addressed throughout the text. It is recommended to revise the conclusion to clearly distinguish between established 5G findings and speculative projections for 6G.

Response 5:

The 6G part was pivoted away from during the writing. We omitted these parts now.

 

Comments 6:

1. The economic feasibility analysis states that the installation cost of the vessel ranges from 70,000 to 100,000 euros. However, it lacks details regarding cost components (e.g., hardware, installation, maintenance) and whether long-term operation and maintenance costs (e.g., software updates, equipment replacement) have been accounted for. Additional clarification is therefore recommended.

Response 6:

The cost breakdown was considered, but since it relies on specific business strategy of the provider, it was eventually omitted. Depending on the business model (service, rental etc) the cost detail varies, but eventual TCO remains roughly the same.

 

Comments 7:

1. The numbering of the charts is inconsistent, with Figure 3 appearing twice. This issue should be revised to ensure proper continuity and accuracy in figure labelling.

Response 7:

This has been resolved in the editing.

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The authors need to address the queries logically. Need to have elaborate explanation with changes in the manuscript simultaneously. The present responses are too short and ambiguous.

1. In the first question, I raised a concern regarding the lack of any formal mathematical modeling, such as the inclusion of a sample raw AIS message. The response indicated that the data was obtained from a pre-processed database. However, the authors could have provided more clarity on the nature of this data, even if no post-processing was required. The methodology could have been supported by a block diagram or a relevant mathematical concept. Additionally, details about the type of information in the pre-processed data; such as latitude, longitude, etc., were missing. The technical explanation in this regard appears inadequate.

2. While the equations are labeled, they do not clearly indicate how the information from the pre-processed AIS database was derived. The Methodology section was expected to be more comprehensive and include mathematical formulations. Furthermore, key terms are missing from equations 1, 2, and 3.

3. The figures still appear to be of low resolution, despite expectations that they would be replaced with improved versions.

4. In some sections, numbered lists are continued from the previous section, creating ambiguity. This is evident in sections 2.1.2 and 2.2, as well as in sections 5.1, 5.2, and 5.3.

5. The Background section is overly descriptive. A tabulated format could improve clarity and focus. Excessive detail in this section tends to detract from the main objective of the paper.

6. The majority of the proposed work focuses on deploying mobile networks at sea as an alternative to satellite communication. However, the emphasis on 5G, a technology subject to generational changes every 10 years, may be too narrow. A broader framing; such as mobile network solutions at sea; may be more appropriate (refer to Comment 7).

 

Author Response

Response 0:

Thank you for additional comments.

Comments 1:

    In the first question, I raised a concern regarding the lack of any formal mathematical modeling, such as the inclusion of a sample raw AIS message. The response indicated that the data was obtained from a pre-processed database. However, the authors could have provided more clarity on the nature of this data, even if no post-processing was required. The methodology could have been supported by a block diagram or a relevant mathematical concept. Additionally, details about the type of information in the pre-processed data; such as latitude, longitude, etc., were missing. The technical explanation in this regard appears inadequate.

Response 1:

The relevant sections has been restructured to better explain the work done and nature of the processed data has been described in a bulleted list. We also provided block diagram describing the workflow.

Comments 2:

 While the equations are labeled, they do not clearly indicate how the information from the pre-processed AIS database was derived. The Methodology section was expected to be more comprehensive and include mathematical formulations. Furthermore, key terms are missing from equations 1, 2, and 3.

Response 2:

Please see updates. Hopefully the improvements in methodology section describe it better now, and makes it easier to follow.

Comments 3:

    The figures still appear to be of low resolution, despite expectations that they would be replaced with improved versions. 

    In some sections, numbered lists are continued from the previous section, creating ambiguity. This is evident in sections 2.1.2 and 2.2, as well as in sections 5.1, 5.2, and 5.3.

Response 3:

This has been fixed now, apparently the issue reappeared due to working with multiple versions of same document.

Comments 4:

    The Background section is overly descriptive. A tabulated format could improve clarity and focus. Excessive detail in this section tends to detract from the main objective of the paper.

Response 4:
We have replaced major part of the section with a table and reduced the amount of detail.

Comments 5:

   The majority of the proposed work focuses on deploying mobile networks at sea as an alternative to satellite communication. However, the emphasis on 5G, a technology subject to generational changes every 10 years, may be too narrow. A broader framing; such as mobile network solutions at sea; may be more appropriate (refer to Comment 7).

Response 5:

This is an issue that we did struggle a lot with internally. The study started deeply rooted in 5G solutions and thus the technological requirements and limitations were based on existing 5G technologies. Therefore we see it is a relevant limitation currently. We have investigated a wider area of technologies (include historical backhaul solution) and also from reviewed literature perspective, but due to latest technical tests covered mainly in 5G domain, we felt that focusing on this area seemed the best in urrent article. On other hand we continue the investigation further, as also additional testings will be addressed soon ( including 5g satellite solutions and distribution algorithms for frequency distribution) and target to address more detailed  approach for even narrower chosen maritime scenarios.

 

Round 3

Reviewer 1 Report

Comments and Suggestions for Authors

The authors have attempted to clarify all the points raised. Can be accepted.