Channel State Information Preprocessing for CSI-based Physical-Layer Authentication Using Reconciliation

TL;DR

The paper tackles CSI-based physical-layer authentication by introducing an adaptive preprocessing step (A-RPCA) that leverages temporal correlations via a TR-PCP-inspired objective. This preprocessing feeds into a reconciliation-based PLA framework using Gaussian-approximation polar codes, enabling reliable cross-time authentication with the Slepian-Wolf decoding setup. The authors provide a convergence analysis for the A-RPCA algorithm and demonstrate substantial gains in CSI correlation, reduced bit-mismatch rates, and near-perfect detection probabilities in both LOS and NLOS scenarios, validated on synthetic data and a Nokia real dataset. Overall, the approach significantly improves robustness and accuracy of CSI-PLA under time-varying conditions and outperforms PCA, RPCA, AE, and ReProCS baselines.

Abstract

This paper introduces an adaptive preprocessing technique to enhance the accuracy of channel state information-based physical layer authentication (CSI-PLA) alleviating CSI variations and inconsistencies in the time domain. To this end, we develop an adaptive robust principal component analysis (A-RPCA) preprocessing method based on robust principal component analysis (RPCA). The performance evaluation is then conducted using a PLA framework based on information reconciliation, in which Gaussian approximation (GA) for Polar codes is leveraged for the design of short codelength Slepian Wolf decoders. Furthermore, an analysis of the proposed A-RPCA methods is carried out. Simulation results show that compared to a baseline scheme without preprocessing and without reconciliation, the proposed A-RPCA method substantially reduces the error probability after reconciliation and also substantially increases the detection probabilities that is also 1 in both line-of-sight (LOS) and non-line-of-sight (NLOS) scenarios. We have compared against state-of the-art preprocessing schemes in both synthetic and real datasets, including principal component analysis (PCA) and robust PCA, autoencoders and the recursive projected compressive sensing (ReProCS) framework and we have validated the superior performance of the proposed approach.

Figures (16)

• Figure 1: Representation of the relationship between CSI dataset at time $t$ and $t+1$ before PCA and after PCA reconstruction: (i) when the correlation is concentrated in top principal components, the preprocessing can improve the correlation and (ii) when the correlation is not concentrated in the largest principal components, the correlation cannot be improved.
• Figure 2: Block diagram of the adaptive preprocessing scheme for CSI-PLA. The green arrow is the path of CSI of Alice $1$ at time $t$ and the pink one is the path of the CSI of Alice $1$ or Alice $2$ during the next time slot $t+1$. The preprocessed channel state information are introduced in the input of the reconciliation block where a quantization is then applied followed by the Slepian-Wolf decoding and the outputs are $\mathbf{r}_1$ and $\mathbf{r}_u$, $u \in \{1, 2\}$.
• Figure 3: A-RPCA preprocessing. The effective CSI (low-rank matrix) at time $t$ is used as helper during next time $t+1$ in the temporal consistency term to perform A-RPCA.
• Figure 4: Heatmap representing the correlation coefficient between the CSI after preprocessing at $t$ and $t+1$: the hypothesis $H_0$ is shown by the yellow and $H_1$ by the red square. The SNR $= 10dB$.
• Figure 5: Bit mismatch rate: $\mathrm{SNR} = 10dB$, $N = 128$, $n = 1$ bit

Theorems & Definitions (9)

• Lemma 1
• proof
• Lemma 2
• proof
• Theorem 1
• proof
• proof
• proof
• proof