jaseg/sampling-mesh-monitor
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Update revision letter

Browse source
This commit is contained in:
jaseg 2025-09-01 13:54:09 +02:00
parent f3f2354c8b
commit 980ed536ab
1 changed files with 17 additions and 15 deletions
Download patch file Download diff file Expand all files Collapse all files

32
paper/revision-letter.tex
View file

@ -4,9 +4,9 @@
\section*{Revision Letter: High Fidelity Security Mesh Monitoring using Low-Cost, Embedded Time Domain Reflectometry}
\vspace*{10mm}
We are grateful for the reviewers' thorough and insightful feedback from both the 2025/4 review cycle. Below, we have
We are grateful for the reviewers' thorough and insightful feedback during the 2025/4 review cycle. Below, we have
compiled a list of questions and comments by the reviewers. We will give detail how we addressed each point in the
present major revision and answer any questions.
present 2026/1 review cycle major revision and answer any questions.
Reviewer A stated,
\begin{displayquote}
@ -14,9 +14,9 @@ Reviewer A stated,
\end{displayquote}
We have adressed Reviewer A's concern in the major revision by reworking all experiments, collecting a large amount of
additional data. In the new experimental section, each experiment has been repeated multiple times, and it is based on a
total of about 1200 additional measurement series. Each measurement series consists of twelve traces and two calibration
traces.
additional data. In the new experimental section, each experiment has been repeated multiple times, and the section is
based on a total of about 1200 additional measurement series. Each measurement series consists of twelve traces and two
calibration traces of 768 points each.
Reviewer A asked,
\begin{displayquote}
@ -26,12 +26,13 @@ Reviewer A asked,
\end{displayquote}
In the major revision, We have adressed Reviewer A's questions in two separate ways. First, we have conducted additional
experiments under a wide range of real-world environmental scenarios including inside a server room. These new results
can now be found in Section 5.3.6 and Figures 14-16. Furthermore, we have improved our experimental design based on
Reviewer A's helpful input to more clearly separate the experiment condition from concurrent environmental effects. To
separate both effects, we took alternating experiment and control measurements, ensuring that environmental effects
affect both groups equally. The experiments shown in Section 5.3.5 and Figure 11, showing performance under
micro-soldering patching attacks, benefitted the most from this improved experimental design.
experiments under a wide range of real-world environmental scenarios including inside a server room where
electromagnetic interference is abundant. These new results can now be found in Section 5.3.6 and Figures 14-16.
Furthermore, we have improved our experimental design based on Reviewer A's helpful input to more clearly separate the
experiment condition from concurrent environmental effects. To separate both effects in runs not focused on
environmental effects, we took alternating experiment and control measurements, ensuring that environmental effects
affect both groups equally. The experiments shown in Section 5.3.5 and Figure 11 where performance under
micro-soldering patching attacks is displayed benefitted the most from this improved experimental design.
Reviewer A asked,
\begin{displayquote}
@ -86,7 +87,8 @@ detail.
During the evaluation of our new measurement data, we noticed PUF-like behavior of our system using our baseline
classifier as mentioned in Section 5.3.1. We believe a closer investigation of this behavior including an entropy
analysis and the construction of an optimized classifier would be an interesting topic for future work. We provide an
outline of this setting in the revised paper in Section 6, ``Future Work''.
outline of this setting in the revised paper in Section 6, ``Future Work''. We expect that once this effect has been
characterized, it can serve as a basis for improved attack detection in a more advanced classifier.
Reviewer C noted,
\begin{displayquote}
@ -96,9 +98,9 @@ Reviewer C noted,
\end{displayquote}
We fully agree with this evaluation and we have provided CAD data, source code, and raw data with the revised paper
during submission. We added anonymized source code and schematics as supplementary material in HotCRP. The raw data set
was too large for direct upload in HotCRP, so we provided it as a (non-anonymous) download link as a comment to the
editors-in-chief.
during submission. We added anonymized source code and schematics as supplementary material to the major revision in
HotCRP. The raw data set was too large for direct upload in HotCRP, so we provided it as a (non-anonymous) download link
as a comment to the editors-in-chief.
In conclusion, we wish to thank the reviewers for their thorough and insightful reviews containing many helpful comments
and interesting questions. We hope we have provided clarity on all points raised, and we believe with the reviewers'