An improved forecasting algorithm for wireless network traffic

With the growing application of wireless networks, the forecasting technologies for wireless network traffic have played a significant role in network management, congestion control and network security. Local Support Vector Machine (LSVM) is an effective method to deal with model for wireless network traffic. To further improve the forecast accuracy and the real-time computing capability of LSVM-DTW-K algorithm we previously proposed based on LSVM, Hannan-Quinn information criterion (HQ) is used to calculate the number of the nearest neighbor points and Symbolic Aggregate Approximation (SAX) is used to symbolic the time series before using Dynamic Time Wrapping (DTW) algorithm to measure the similarity between two points.