改进粒子群算法优化BP网络实现流量预测

The Realization of Traffic Prediction by Improving Particle Swarm Optimization and Optimizing BP Network

近年来,随着网络技术的不断发展,网络传输业务也在不断增加,网络业务流量数据与日俱增。飞速增长的流量数据给网络结构和信息传输带来深刻的影响,大量的网络数据在有限的空间内进行传输,网络拥塞现象难以避免,造成网络传输时延增大、传输速率降低。因此,为了更有效地进行网络管理,改善网络传输性能,有效预测业务流量数据的变化趋势,实行具体、全面的网络管理是十分必要的。该文针对网络流量表现出的自相似性和可预测性,提出了改进PSO-BP算法,并建立其预测模型。结果显示,在网络流量预测领域内,与常规BP网络流量预测算法相比,引入改进PSO-BP算法的神经网络预测模型表现出更为优异的性能。

In recent years,with the continuous development of network technology,network transmission services are also increasing,and network traffic data is increasing day by day.The traffic data of rapid growth has a profound impact on the network structure and information transmission,a large amount of network data is transmitted in a limited space,and network congestion is unavoidable,resulting in the consequence of the increase of network transmission delay and the decrease of transmission rate.Therefore,in order to manage the network more effectively,improve the network transmission performance,and effectively predict the change trend of traffic data,it is necessary to implement specific and comprehensive network management.Aiming at the self-similarity and predictability of network traffic,this paper proposes an improved PSO-BP algorithm and establishes its prediction model.Results show that in the field of network traffic prediction,compared with the conventional BP network traffic prediction algorithm,the neural network prediction model with the improved PSO-BP algorithm shows better performance.