Field survey of wireless ISM-band channel properties for substation applications

During this field survey, we measured and recorded a few quality parameters of wireless communication in a substation switchyard. A microprocessor-based measurement system was used for data collection and analysis. We investigated long-term noise variation in this specific environment. Based on our measurement and post-processing analysis we conclude that the so-called classic/Bayesian assumption of existing noise distributions appear not to be a suitable all-round model for analysis. Variations in bandwidth occupancy patterns of other wireless devices sharing the same frequency and many other factors necessitate updated measurements and post processing. We noticed dominant underlying structure in noise profiles, which calls for a comprehensive time series analysis. Given the stationarity of the data set and the Wold´s theorem, an auto regressive moving average (ARMA) model can be found for the data set with similar behaviors.