RIS-CLA: Reviving CSI-Based Continuous Location Authentication With Reconfigurable Intelligent Surfaces

Continuous Location Authentication (CLA) enables seamless verification of the physical locations of networked devices. While wireless Channel State Information (CSI)based CLA schemes are attractive due to their passive and infrastructure-compatible design, recent studies have exposed critical vulnerabilities to CSI-guessing attacks, limiting their applicability in high-security environments. We propose RISCLA, a novel system that leverages Reconfigurable Intelligent Surfaces (RISs) to enable secure and autonomous CSI-based CLA. By dynamically reconfiguring the wireless propagation environment through RIS control codewords, RIS-CLA induces unpredictable, device-specific CSI variations that adversaries cannot easily guess or replicate. The system addresses three key challenges-codeword ambiguity, robustness to channel dynamics, and multi-device scalability-through a suite of techniques including codeword optimization, contrastive-learning-based classification, explainable-AI-guided model finetuning, and a joint multi-device authentication framework. Extensive prototype-based experiments demonstrate that RISCLA effectively mitigates CSI-guessing attacks and achieves robust CLA performance under diverse conditions.