基于卡尔曼滤波和小波的网络流量预测算法研究

Application of Kalman Filter and Wavelet in Traffic Prediction

流量预测是流量工程,拥塞控制和网络管理的核心问题。该文针对网络流量的特点,将卡尔曼滤波和小波分析混合的预测算法引入到网络流量预测领域中,对其进行了理论证明。仿真结果表明,该算法与传统的算法相比,具有较高的预测精度和较好的实时性与广谱性。

Traffic prediction is the core of network quality of service problems, such as traffic engineering and congestion control, etc. According to the characters of traffic, a novel network traffic prediction algorithm in which Kalman filter and wavelet are mixed is presented and proved abstractly. The simulation results show that the proposed algorithm can guarantee higher precision and better real-time processing compared with traditional algorithm.