Research On AOS Frame Generation Technology Based On Self-similar Traffic Prediction

With the increasing complexity of the space network system and the transmission of data,people's requirements for space communication and data transmission and processing are also constantly improving.The performance of the network system directly affects the correctness and reliability of data transmission.The study of a series of spatial network protocols for data has become particularly important.A large number of literatures and research results show that network traffic has self-similarity,which provides a new idea for in-depth study of self-similar traffic prediction and multiplexing of spatial networks.The self-similar traffic prediction module can be introduced into the multiplexing model.Further improve the multiplex optimization system model.This paper studies the frame generation technology of Advanced Orbiting System(AOS)based on self-similar traffic prediction,and optimizes the AOS multiplexing model based on the predictability of self-similarity.First,the characteristics of self-similar traffic and the frame generation algorithm are briefly introduced,and a new AOS frame generation algorithm based on self-similar traffic is proposed.The algorithm determines the adaptive frame generation threshold and controls the frame generation time based on the comprehensive index of average packet delay and frame multiplexing efficiency,and realizes the intelligence of the adaptive framing algorithm.Finally,the Matlab platform is used to simulate and verify the newly improved AOS adaptive frame generation algorithm according to the theoretical formula.The simulation results show that this algorithm can optimize the performance of the multiplexing model.Secondly,research on self-similar traffic prediction technology based on intelligent algorithms.This article uses a non-linear prediction model.It uses a neural network model to predict traffic.The prediction principle of this model is relativelyeasy to implement and the prediction speed is fast.On this basis,the neural network prediction model based on wavelet function is studied,the self-similarity of traffic is predicted,and the prediction model is simulated and verified.Finally,the AOS frame generation algorithm based on self-similar traffic prediction is studied.The predicted self-similar traffic is introduced into the AOS frame generation module,and an improved adaptive frame generation algorithm and a scheduling algorithm are combined to form an AOS frame generation algorithm based on traffic prediction.In this paper,the above algorithm is verified by overall simulation,and the impact on the system with and without the prediction module is compared.The final results show that the system with the prediction algorithm is significantly better than the system without the prediction algorithm in improving the average packet delay,system remaining capacity,and frame loss rate.