2018/09/24 by Simone Grimaldi, Grimaldi, Simone, Aamir Mahmood +3
Computer Science · Engineering · #Bluetooth and Wireless Communication Technologies #FOS: Electrical engineering #Millimeter-Wave Propagation and Modeling #Signal Processing (eess.SP) #Wireless Networks and Protocols #electronic engineering #information engineering
paper · pdf · doi:10.48550/arxiv.1809.10085
openalex publication_date 2018/09/24 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
Energy sampling-based interference detection and identification (IDI) methods\ncollide with the limitations of commercial off-the-shelf (COTS) IoT hardware.\nMoreover, long sensing times, complexity and inability to track concurrent\ninterference strongly inhibit their applicability in most IoT deployments.\nMotivated by the increasing need for on-device IDI for wireless coexistence, we\ndevelop a lightweight and efficient method targeting interference\nidentification already at the level of single interference bursts. Our method\nexploits real-time extraction of envelope and model-aided spectral features,\nspecifically designed considering the physical properties of signals captured\nwith COTS hardware. We adopt manifold supervised-learning (SL) classifiers\nensuring suitable performance and complexity trade-off for IoT platforms with\ndifferent computational capabilities. The proposed IDI method is capable of\nreal-time identification of IEEE 802.11b/g/n, 802.15.4, 802.15.1 and Bluetooth\nLow Energy wireless standards, enabling isolation and extraction of\nstandard-specific traffic statistics even in the case of heavy concurrent\ninterference. We perform an experimental study in real environments with\nheterogeneous interference scenarios, showing 90%-97% burst identification\naccuracy. Meanwhile, the lightweight SL methods, running online on wireless\nsensor networks-COTS hardware, ensure sub-ms identification time and limited\nperformance gap from machine-learning approaches.\n