Current research has found that actual network traffic has obvious fractal characteristics. The multi-fractal characteristics of traffic have a very important impact on network performance. It is necessary to establish an actual network business model based on multi-fractal characteristics that can simultaneously predict long-correlation and short-correlation characteristics. The AR, ARMA and other models can predict short-correlation data well but have low prediction accuracy for long-correlation data. In combination with wavelet transform, the actual data correlation can be removed to establish a new prediction model, which also has relatively high prediction accuracy for long-correlation data. The improved model overcomes the disadvantage of the FARIMA model\'s relatively large amount of calculation and maintains the simplicity of the algorithm. Keywords: multi-fractal; long-correlation; wavelet; prediction
You Might Like
Recommended ContentMore
Open source project More
Popular Components
Searched by Users
Just Take a LookMore
Trending Downloads
Trending ArticlesMore