| With the rapid development of the mobile Internet and the rise of The Internet of Things(IOT),the network transmission scenarios are becoming more and more diversified,and the transmission contents are becoming more and more complex.Faced with the complex network environment and the high requirements of users on the communication transmission,many protocols have emerged as the times require.Among them,Quick UDP Internet Connection(QUIC)has been studied due to its flexibility,clear framework,easy deployment,no head-of-line blocking and other advantages.However,when facing the communication requirements of audio,video,and VR and other high latency sensitive services,the integrated converged network that only relies on the deployment of QUIC protocol still has some problems,such as failing to meet users’ good network experience and continuous high-speed network loading in the state of emergency.As a result,the Deadline-aware Transport Protocol(DTP),which is based on the extension of the QUIC protocol,proposes a single-path delivery deadline packet scheduling,but it does not sufficiently consider the use of multi-link cooperation to improve transmission efficiency.Therefore,this thesis designs single-path adaptive migration and multi-path scheduling methods to improve network resource utilization,network transmission efficiency,and transmission reliability.The main work of this thesis is as follows:Firstly,this thesis proposes Proactive Connection Migration(PCM),an adaptive migration mechanism based on QUIC,which combines response delay and jitter to complete path exploration from multiple different dimensions.PCM is a highly compatible and cross-platform service mechanism.By using adaptive connection migration,it improves the network user experience,ensures the rational utilization of link transmission resources,and fills the gap of QUIC optimal performance through path exploration in a multi-network environment.The thesis focuses on the design overview and model of PCM,and introduces the scheme and implementation of the scheme in detail.Secondly,this thesis proposes a multi-path cooperative transmission scheduling algorithm Tetris based on DTP,and expounds its applicable scenarios,the design and implementation details of different modules in the algorithm.Tetris considers parameters such as deadline,block priority,creation time,and block size and designs a near-optimal scheduling based on deadline awareness by means of mathematical modeling,ensuring that as many data blocks as possible are delivered before the delivery deadline,effectively balancing the relationship between block priority and block delivery deadline,making full use of link resources,and improving the reliability and effectiveness of concurrent transmission.Finally,based on the above design,the algorithm strategy is built and deployed on the programmable prototype system,the system function and performance of the two mechanisms are verified.The results show that the PCM mechanism has high link switching accuracy,fast link convergence speed and can efficiently select the optimal link in various integrated converged network scenarios.In PCM,the mean response time of pessimistic Path Selection Algorithm(PSA)decreased by 51.63%,and the variance decreased by 23.47% compared to polling strategy;the mean response time of optimistic PSA decreased by 43.17%,and the variance decreased by-7% compared to polling strategy,which can effectively improve the transmission efficiency.The multi-path coordinated transmission scheduling algorithm Tetris is less affected by link fluctuations,whose performance can be improved by 1-2.5 times compared to other algorithms in various transmission scenarios.Its transmission completion time,cumulative distribution function,and Qo E have better performance in different network scenarios,which have stronger practical significance for network link communication transmission scheduling. |