Alogirithm Design And Analysis Of GCC Handover Detection And Rate Recovery For Low Earth Orbit Satellite Networks

Low earth orbit satellite networks have developed rapidly in recent years and will become an important part of the next-generation Internet.In order to ensure users’experience,the application should be optimized for the satellite network scenario.Real-time audio and video applications represented by Web Real-Time Communications(WebRTC)have attracted wide attention in satellite networks.We test the performance of WebRTC in an emulated standard-topology satellite network platform and the results show that due to the inability to distinguish between the satellite-ground link handover and network congestion,Google Congestion Control,which is the default congestion control algorithm for WebRTC,cannot adopt an appropriate rate recovery strategy for handover,resulting in an extended period of video frame stalling.Therefore,the ability to detect handover and perform fast recovery after handover is the key to improve the QoS of WebRTC application in satellite networks.Based on the study of different congestion control algorithms and low earth orbit satellite networks,we focus on the detection of satellite-earth link handover and fast rate recovery.For the problem of handover detection,we propose a handover detection algorithm based on the satellite remaining service duration.We use outlier detection to filter the extremely high loss rate and judge the occurrence of handover.To improve the accuracy of detection,we provide the detection window by computing the remaining service time of the satellite and limit the time period in which the detection may occur.To solve the problem that GCC cannot recover quickly after handover,we propose a fast handover recovery algorithm based on historical network information.Inspired by the TCP fast retransmission,we use time series prediction algorithm to provide an initial sending rate for GCC.Then,based on the historical packet loss rate,we provide an adaptive coefficient for GCC’s rate control formula.Finally,a transmission simulation platform for low earth orbit satellite network is implemented,which provides a verification environment for the above two algorithms.The experimental results show that our handover detection algorithm can guarantee accurate recognition of handover.Compared to ordinary GCC,the proposed rate recovery algorithm can significantly reduce the time to recovery and improve the video fluency of WebRTC.