2201.06017
Submodularity-based False Data Injection Attack Scheme in Multi-agent Dynamical Systems
Xiaoyu Luo, Chengcheng Zhao, Chongrong Fang, Jianping He
wronghigh confidenceCounterexample detected
- Category
- math.DS
- Journal tier
- Specialist/Solid
- Processed
- Sep 28, 2025, 12:56 AM
- arXiv Links
- Abstract ↗PDF ↗
Audit review
The paper’s unconditional submodularity claim for the time-variant attack case (Theorem 3) relies on two invalid steps: (i) a monotonicity argument that assumes the norm of a sum is always at least the norm of a summand, and (ii) a submodularity proof that sets all cross terms (componentwise products) to zero without justification. Both steps are mathematically incorrect. The candidate model exhibits these flaws and provides explicit counterexamples within the paper’s assumptions, showing submodularity (and even monotonicity) can fail.
Referee report (LaTeX)
\textbf{Recommendation:} reject
\textbf{Journal Tier:} specialist/solid
\textbf{Justification:}
The central theoretical claim—that the convergence error is always submodular and monotone in the time-variant case, independent of system details—is not supported. The proof contains invalid steps (misuse of norm inequalities; unjustified vanishing of cross terms), and simulation results in the paper itself hint at counterexamples. Without a corrected, scoped theorem and rigorous assumptions, the paper’s main contribution does not hold.