Channel Impulse Response-based Physical Layer Authentication in a Diffusion-based Molecular Communication System

Referee: [Abstract] Abstract: the claim that 'simulation results highlight the great challenge posed by impersonation attack--i.e., it is not possible to simultaneously minimize the two error probabilities' is load-bearing for the paper's conclusion yet supplies no quantitative results, error probability values, simulation parameters (diffusion coefficients, node distances, L, slot timing), error bars, or hypothesis-test threshold details. This prevents verification of the reported tradeoff.

Authors: We agree that the abstract would benefit from explicit quantitative details to support the central claim. In the revised manuscript we will expand the abstract to report key simulation parameters (diffusion coefficient, node distances, L, slot duration), representative false-alarm and miss-detection probabilities obtained from the Monte-Carlo trials, and the decision threshold employed in the hypothesis test. This will allow readers to verify the reported error-probability tradeoff directly from the abstract. revision: yes

Referee: [—] Description of the binary hypothesis test (using Alice's measured 3D CIR as ground truth): the manuscript provides no analysis or results demonstrating that the CIR is sufficiently unique across nano-nodes or stable enough under diffusion dynamics that Eve cannot produce a statistically close response to pass the test. The noted error tradeoff alone does not establish the required discriminative power.

Authors: The simulations already embed the hypothesis test that compares the received 3-D CIR vector against Alice’s measured ground-truth vector; the resulting error tradeoff arises precisely because diffusion and finite slot timing limit how distinguishable two transmitters can be. Nevertheless, we acknowledge that an explicit discussion of uniqueness would strengthen the presentation. In revision we will add a short subsection that (i) derives the dependence of each CIR tap on transmitter–receiver distance and diffusion coefficient, (ii) shows via additional simulation curves that an Eve located at a different coordinate produces a statistically distinguishable CIR vector, and (iii) quantifies the residual overlap that forces the observed false-alarm / miss-detection tradeoff. These additions will directly address the discriminative-power concern while remaining within the scope of the existing simulation framework. revision: yes