Experimental Evaluation of A Lightweight RSS-Based PLA Scheme in Multi-Node Multi-Cell Mesh Networks

In infrastructure-less Internet of Things (IoT) mesh networks, node authentication is a critical challenge. The situation becomes even more complex when considering the utilization of resource-constrained and low-end devices, as this restricts the applicability of existing cryptographic solutions. Physical layer authentication (PLA) has recently emerged as a complementary solution to enable a secure authentication process for infrastructure-less IoT deployments. Consequently, this research paper aims to empirically investigate the utilization of a PHY layer-based feature, namely received signal strength (RSS), for establishing a robust PLA in multi-node (MN) multi-cell (MC) wireless mesh networks. First, we develop a low-complex yet secure and efficient RSS-based PLA framework to execute continuous authentication for multiple legitimate nodes from different cells. Then, through extensive experiments, we evaluate the performance of the proposed scheme in terms of authentication rate, the detection rate of illegitimate nodes, and variations in the reference RSS values. Our evaluations consider different numbers of nodes, their respective locations, and relative distances concerning the receiver node (authenticator node). Furthermore, through a comprehensive analysis of the experimental data, we demonstrate the impact of the RSS threshold on the overall system performance. The obtained results demonstrate that the authentication rate is directly proportional to the RSS threshold while the reverse is correct for the detection rate.