In this paper, a deep learning (DL)-based physical (PHY) layer authentication framework is\nproposed to enhance the security of industrial wireless sensor networks (IWSNs). Three algorithms,\nthe deep neural network (DNN)-based sensor nodesâ?? authentication method, the convolutional neural\nnetwork (CNN)-based sensor nodesâ?? authentication method, and the convolution preprocessing\nneural network (CPNN)-based sensor nodesâ?? authentication method, have been adopted to implement\nthe PHY-layer authentication in IWSNs. Among them, the improved CPNN-based algorithm requires\nfew computing resources and has extremely low latency, which enable a lightweight multi-node\nPHY-layer authentication. The adaptive moment estimation (Adam) accelerated gradient algorithm\nand minibatch skill are used to accelerate the training of the neural networks. Simulations are\nperformed to evaluate the performance of each algorithm and a brief analysis of the application\nscenarios for each algorithm is discussed. Moreover, the experiments have been performed with\nuniversal software radio peripherals (USRPs) to evaluate the authentication performance of the\nproposed algorithms. Due to the trainings being performed on the edge sides, the proposed method\ncan implement a lightweight authentication for the sensor nodes under the edge computing (EC)\nsystem in IWSNs.
Loading....