Classification of Radio Channel disturbances for industrial wireless sensor networks

The reliability of data transmission in Wireless Sensor Networks (WSN) is always an issue in harsh industrial environments and sets specific challenges for performance optimization. Short-term signal disturbances, in the form of multipath fading and destructive radio interference, are of major concern due to signal path conditions (large concrete and metal surfaces), related line-of-sight (LOS) changes (incoming and outgoing trucks, moving around forklifts and workers), and radio frequency interferences. In this paper we introduce the classification procedure where results of Probability Density Function (PDF) analysis and Spectrogram Analysis are combined to classify the measured radio channel disturbances to a set of predefined disturbance classes. The PDF and Spectrogram analysis methods are used for analyzing the statistical properties of the received signal magnitudes. Tracking the changes of the PDF clearly contributes to recognize and characterize the temporal changes, especially LOS changes in the radio link environment. The spectrogram analysis provides additional information of the radio interferences on coand adjacent channels. © 2016 Elsevier B.V. All rights reserved.