Off-Cloud Anchor Sharing Framework for Multi-User and Multi-Platform Mixed Reality Applications
, Oscar Blanco-Novoa 1,2, Paula Fraga-Lamas 1,2,* and Tiago M. Fernández-Caramés 1,2
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsIn the manuscript, the author proposed an innovative framework called "Off-cloud Anchor Sharing Framework" to solve the problem of spatial anchor sharing in multi-user, multi-platform MR applications. The author's experiment had the following problems: 1. Insufficient sample size. The experiment was repeated only 5 times for each scenario, and the statistical significance was insufficient, making it difficult to fully reflect the stability of the system under different loads. 2. The experiment was only conducted in a local area network environment, and the performance in public network or weak network (such as high latency, packet loss) scenarios was not tested. In actual applications, network fluctuations may cause anchor synchronization failures or a sharp increase in latency. 3. Insufficient cross-platform testing. Although it claims to support multiple platforms, the experiment only involves HoloLens 2 and desktop terminals, and the compatibility and performance differences of mobile devices such as iOS and Android have not been verified.
Author Response
The authors would like to thank the reviewers for their valuable comments, which have certainly helped us to improve the manuscript. Please find attached our detailed responses to the comments.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsThis manuscript proposes an off-cloud spatial anchor sharing framework for multi-user, cross-platform mixed reality (MR) experiences. The motivation is clear: with the deprecation of Azure Spatial Anchors, developers require alternative solutions that are self-hosted, cost-effective, and privacy-conscious. The authors implement a LAN-based sharing mechanism using Mirror networking and local anchor serialization, and present timing benchmarks from two HoloLens 2 devices using varying anchor sizes. However, the overall presentation of the work lacks the clarity, depth, and rigor required for publication at this stage. 1）From a logical perspective, the manuscript does not clearly distinguish its novelty. While it claims to be an off-cloud alternative, it offers little structured comparison to other open-source or local anchor-sharing solutions such as Mozilla Hubs or ArUco marker systems. Without such comparative analysis, it is difficult to determine the system’s unique contribution. Additionally, several key claims—such as real-time performance, cost-efficiency, and enhanced privacy—are not sufficiently supported by empirical data or theoretical modeling. 2）the maniscript suffers from inconsistent language quality. There are repeated typographical errors (e.g., “syncronization”), redundant phrasing (e.g., “makes use of”), and inconsistencies in formatting between abstract sections. 3)The experimental section is too narrow to substantiate the system’s scalability and robustness. Only two devices are tested under ideal network conditions. There is no evaluation under adverse bandwidth scenarios, no measurement of synchronization accuracy, and no stress tests with more than two users or heterogeneous device types. Key metrics such as CPU usage, energy consumption, and perceptual drift are absent. To truly demonstrate system viability, the authors should expand the testing to include more devices, simulate real-world networking conditions, and report perceptual accuracy metrics.
Author Response
The authors would like to thank the reviewers for their valuable comments, which have certainly helped us to improve the manuscript. Please find attached our detailed responses to the comments.
Author Response File: Author Response.pdf
Reviewer 3 Report
Comments and Suggestions for Authors- The experiment only shows the performance data of the self-developed framework and does not compare it with similar solutions, making it difficult to quantitatively evaluate its advantages.
- The research focuses on technical performance but does not assess the user experience in actual use, such as the delay perception during multi-user collaboration and the smoothness of device interaction.
- The paper emphasizes local area network (LAN) deployment but does not discuss the expansion scheme in wide area network (WAN) scenarios, and future work prospects can be added.
Author Response
The authors would like to thank the reviewers for their valuable comments, which have certainly helped us to improve the manuscript. Please find attached our detailed responses to the comments.
Author Response File: Author Response.pdf
Reviewer 4 Report
Comments and Suggestions for AuthorsThe paper proposes a space anchor synchronization mechanism that does not rely on third-party cloud services. It enables real-time collaboration among multiple devices through local storage and efficient communication protocols, resolving cross-platform compatibility and multi-user synchronization issues and mitigating the risks associated with the suspension of cloud services. A multi-user collaborative game case was developed using Unity, Mirror, and MRTK, combined with IoT modules to enhance interaction between the virtual and real worlds. The open-source code was provided to improve reproducibility. A segmented transmission protocol and an automatic discovery module were designed, reducing the transmission time of small anchors to 20 seconds and that of complex scenes to 90 seconds. The cost of repeated transmission was also reduced through the use of ontology storage. However, the paper has the following shortcomings:
1. The abstract section mentions "using the Unity, Mirror, and MRTK development framework." However, it does not describe the core technical architecture of the framework (such as the layered module design, anchor synchronization algorithm logic, etc.). The description of the research method is too general. It does not highlight the innovation of the paper, such as the improvement of collaboration autonomy and continuity by the "non-cloud" architecture. The "20 seconds for short anchor transmission time" in the abstract is inconsistent with the 18.095 seconds in the experimental data.
2. The reference literature does not provide a sufficient description of the "non-cloud anchor sharing" field, only mentioning blockchain-assisted solutions.
3. The technical limitations of the local solution (such as differences in device computing power and local network bandwidth restrictions) were not analyzed, and the argument for non-cloud demand derived from the retirement of Azure Spatial Anchors is insufficient.
4. In Figure 1, the "Mirror UI Discovery" does not match the "NetworkDiscovery UI" module name described in the text.
5. In Section 3.3.2, the "local anchor storage format" and "anchor conflict resolution mechanism" of the anchor synchronization algorithm were not described. Please add them.
6. When using a Network Transmitter to segment and transmit anchor data, no optimal packet length calculation (such as a segmentation strategy based on the Maximum Transmission Unit, or MTU) or delay formula was provided. Describing it as "divided into 4096-byte packets" is not rigorous enough.
7. In the Bluetooth automatic discovery mechanism, the device allowlist matching rules (such as UUID filtering logic) or reconnection strategy were not explained. Please explain how the hardware interaction process is implemented.
8. In Section 3.3, it is suggested to add reusable general components (such as an abstract anchor manager interface).
9. How to dynamically load platform-specific resources (such as MRTK input for HoloLens vs. desktop keyboard control)?
10. Attributing the 90-second room-level anchor transmission time to "bandwidth and device performance" is inaccurate and not rigorous.
11. It is recommended to provide specific technical details, supplement theoretical modeling, conduct comparative experiments, and enhance cross-platform adaptation logic.
Author Response
The authors would like to thank the reviewers for their valuable comments, which have certainly helped us to improve the manuscript. Please find attached our detailed responses to the comments.
Author Response File: Author Response.pdf
Round 2
Reviewer 2 Report
Comments and Suggestions for AuthorsThe current version has solved my previous problem, but there are still some grammar errors.
Author Response
The authors would like to thank the reviewers for their valuable comments, which have certainly helped us to improve the manuscript. Please find attached our detailed responses to the comments. In order to ease the labor of the reviewers we have colored in red the differences with the previous version of the article.
Author Response File: Author Response.pdf
Reviewer 3 Report
Comments and Suggestions for AuthorsThe issues raised have been addressed.
Author Response
The authors would like to thank the reviewers for their valuable comments, which have certainly helped us to improve the manuscript.
Author Response File: Author Response.pdf
Reviewer 4 Report
Comments and Suggestions for AuthorsAll my comments have been properly accommodated.
Author Response
The authors would like to thank the reviewers for their valuable comments, which have certainly helped us to improve the manuscript.
Author Response File: Author Response.pdf
