Real-Time Measurement Method of Available Bandwidth Based on Kalman Filter

摘要

The available bandwidth measurement is of great significance for the optimization of media service transmission and quality of experience. Bandwidth measurement is divided into active measurement and passive measurement according to the measurement mode. Currently, the available bandwidth measurement techniques are generally based on the Probe Gap Model or Probe Rate Model. The Probe Rate Model which has a great impact on the network is less considered. The shortcoming of Probe Gap Model is it needs to know bottleneck bandwidth first and a Kalman Filter can be used to solve this problem. The Kalman filter has the characteristics of being able to effectively filter out the noise by combining the predicted value and the measured value. Firstly, the exact value of the bottleneck bandwidth is calculated, and then the Probe Gap Model is applied. Finally, real-time measurement is implemented in the Linux environment and analysis of the experiment shows that the results are more accurate.