Abstract:The number of Internet of Things (IoT) deployments is expected to reach 75.4 billion by 2025. Roughly 70% of all IoT devices employ weak or no encryption; thus, putting them and their connected infrastructure at risk of attack by devices that are wrongly authenticated or not authenticated at all. A physical layer security approach -- known as Specific Emitter Identification (SEI) -- has been proposed and is being pursued as a viable IoT security mechanism. SEI is advantageous because it is a passive technique that exploits inherent and distinct features that are unintentionally added to the signal by the IoT Radio Frequency (RF) front-end. SEI's passive exploitation of unintentional signal features removes any need to modify the IoT device, which makes it ideal for existing and future IoT deployments. Despite the amount of SEI research conducted, some challenges must be addressed to make SEI a viable IoT security approach. One challenge is the extraction of SEI features from signals collected under multipath fading conditions. Multipath corrupts the inherent SEI features that are used to discriminate one IoT device from another; thus, degrading authentication performance and increasing the chance of attack. This work presents two semi-supervised Deep Learning (DL) equalization approaches and compares their performance with the current state of the art. The two approaches are the Conditional Generative Adversarial Network (CGAN) and Joint Convolutional Auto-Encoder and Convolutional Neural Network (JCAECNN). Both approaches learn the channel distribution to enable multipath correction while simultaneously preserving the SEI exploited features. CGAN and JCAECNN performance is assessed using a Rayleigh fading channel under degrading SNR, up to thirty-two IoT devices, and two publicly available signal sets. The JCAECNN improves SEI performance by 10% beyond that of the current state of the art.

Real-time Interference Identification via Supervised Learning: Embedding Coexistence Awareness in IoT Devices

EIM: Efficient Online Interference Measurement in Wireless Sensor Networks

A Point-to-Point Interference Measurement Approach for Large-Scale Wireless Sensor Networks.

Deep Learning for Interference Identification: Band, Training SNR, and Sample Selection

Intelligent Spectrum Learning for Wireless Networks with Reconfigurable Intelligent Surfaces.

Reconfigurable Intelligent Surface Enhanced Massive IoT Systems with Nonlinear Measurements

HEIR: Heterogeneous interference recognition for wireless sensor networks

Leveraging Machine Learning for Accurate IoT Device Identification in Dynamic Wireless Contexts

Machine Learning Methods for Device Identification Using Wireless Fingerprinting

Distributed Unsupervised Learning for Interference Management in Integrated Sensing and Communication Systems

Interference Resilient Duty Cycling for Sensor Networks under Co-Existing Environments

Leveraging Deep Learning for IoT Transceiver Identification

Interference Recognition for Fog Enabled IoT Architecture using a Novel Tree-based Method

Machine Learning Based Iterative Detection and Multi-Interference Cancellation for Cognitive IoT

LUCID: Receiver-aware Model-based Data Communication for Low-power Wireless Networks

Performing Joint Learning for Passive Intrusion Detection in Pervasive Wireless Environments.

Improving RF-DNA Fingerprinting Performance in an Indoor Multipath Environment Using Semi-Supervised Learning

Decentralized Spectrum Learning for IoT Wireless Networks Collision Mitigation

Learning-Based Iterative Interference Cancellation For Cognitive Internet Of Things

Interference Suppression Using Deep Learning: Current Approaches and Open Challenges