Jitter Buffer Control Algorithm and Simulation Based on Network Traffic Prediction

摘要

The jitter buffer size is one of the important factors that affect the voice quality of Voice over Internet Protocol (VoIP). In order to reduce buffer delay and packet loss and improve the speech quality of VoIP, an adaptive jitter buffer control algorithm based on network traffic prediction is proposed. Among them, the particle swarm optimization algorithm is used to optimize the weight and threshold of Elman neural network so that it avoids falling into local minimum and improves the accuracy of network traffic prediction. According to the change law of network traffic, the jitter buffer control algorithm under the autoregressive model is proposed. Through the improved stochastic midpoint placement algorithm, a network business traffic prediction model with sudden and self-similarity is established. Then the buffer size is set according to the predicted value of network traffic, and it is continuously improved in use to improve the accuracy of the buffer setting. The simulation results show that the MOS value of the algorithm is high, which improves the voice quality of the network telephone.