Measurement And Analysis Of The Electromagnetic Noise For Industrial Internet Of Things

With the strategic plans of "Industry 4.0","Horizon 2020" and "Made in China 2025”,the industrial manufacturing is moving towards automation and intellectualization step by step.The application of information equipment in the traditional manufacturing industry has become an inevitable requirement to realize the intelligent manufacturing.Industrial Internet of Things Technology(IIOT)can greatly improve the factory’s production efficiency and quality due to low cost and simple deployment.But the electromagnetic noise emitted by mechanical equipment such as frequency converters,voltage regulators and welding machines will affect the IIOT system with low-power wireless sensor chips as the core.Therefore,this paper researches the electromagnetic noise in an automobile factory.The electromagnetic noise in welding workshop and office area of the factory is measured by a spectrum analyzer and a log-periodic antenna from time-domain and frequency-domain.The antenna is polarized vertically and horizontally.We obtain the spectrum occupancy in 300MHz～3GHz bands from the frequency-domain measurement and distinguish the electromagnetic noise from communication signals.We also select typical frequency bands for analyzing and give the spectrum planning scheme for IIOT.According to the measurement results,we can conclude that the noise frequency mainly distributes in 300MHz～1GHz bands and the noise power is mainly between-125dBm and-115dBm.The truncated Laplace distribution and the improved bell shaped spectrum can be used to characterize the frequency spectrum of the electromagnetic noise.In the time-domain measurement,the noise data of 315MHz,779MHz and 916MHz are collected for researching the tine-varying characteristics of the electromagnetic Noise.We also get four parameters of Amplitude Probability Distribution(APD),Pulse Duration Distribution(PDD),Pulse Separation Distribution(PSD)and Noise Amplitude Distribution(NAD).According to the time-domain data,we use CHMM to model the electromagnetic noise.Comparing time-domain statistical characteristics of measured noise and simulated noise generated by CHMM,we can conclude that the Continuous Hidden Markov Model(CHMM)can be used to model the time-varying characteristics of the electromagnetic noise.