Research On Self-similarity Of Traffic In Wireless Sensor Networks For Smart Grid

Wireless sensor networks have been applied to smart grid, which have greatly accelerated the intelligent process of the smart grid. The advantages of wireless sensor networks will be used in more fields for smart grid. However, we also need to slove and avoid the disadvantage of it. The traffic of wireless sensor networks in smart grid is first analyzed in this paper. It is necessary to research on it because it offers theoretical basis for its further usage in smart grid.First, the non-linear characteristic of traffic in wireless sensor networks for smart grid is certified, and we could use fractal theory to discuss it. Then, monofractal theory is used to verify the self-similarity and long range dependence of the traffic. Here we calculate the Hurst values based on different arithmetics, different sample frequencies and different time solts. we also calculate the fractal dimension of the time series. At last, multifractal theory is used here to discuss the multifractal characteristics of the traffic in wireless sensor networks for smart grid. Here MFDFA and multifractal spectrum are analyzed to describe the multifractal characteristics, which depict the self-similarity and non-linear dynamics from the perspective of details and locality.The characteristics of the traffic are analyzed here based on fractal theory, which include the global features, the relationships between the whole and the parts, and the locality features. It provides theory basis for the research on traffic modeling and fault detection.