基于业务流量预测的AOS自适应帧生成算法

AOS Adaptive Frame Generation Algorithm Based on traffic Prediction

随着研究的深入,复杂空间系统业务数据流的自相似性逐渐被认识,传统的等时帧生成算法及高效率生成算法越来越难以适应空间系统业务流量的高突发性和高复杂性;基于此,提出了一种基于小波神经网络业务流量预测的自适应帧生成算法,在满足一定延时约束条件下,根据业务流量预测结果,自适应调整成帧时刻;帧复用效率相比等时帧生成算法有显著优势,同时还避免了高效率帧生成算法中存在的帧延时过长的问题。

With the in-depth study, the self similarity of complex spatial data system is gradually recognized. Traditional time frame generation algorithm and efficient frame generation algorithm are more and more difficult to adapt to the high burst and high complexity of space traffic. This paper presents an adaptive frame generation algorithm based on wavelet neural network traffic prediction. Under the con- dition of certain delay constraint, the adaptive frame generation time can be adjusted according to the prediction results of traffic flow. Corn pared with the time frame generation algorithm, the frame multiplexing efficiency of this algorithm has a significant advantage, and it also a voids the problem of long frame delay.