A real-time multimedia streaming transmission control mechanism based on edge cloud computing and opportunistic approximation optimization

摘要

We address the problem of real-time transmission of multimedia streams in distributed heterogeneous networks. The effect of this problem directly affects the execution efficiency, real-time and reliability of the network system. Firstly, for distributed multimedia data, the heterogeneity of edge cloud devices is considered. Through in-depth analysis of the dependency between multimedia packets and video frames, the edge cloud and its dependent directed acyclic graph are established, and the edge cloud computing model is set up to execute cost, real-time and dependability. Secondly, based on the deadline features of multimedia real-time applications, the optimal solution of the edge cloud sets is established. The establishment basis is from the maximum satisfiability problem and the search for the best edge cloud with the execution cost and execution time of the multimedia streaming real-time communication application. According to the above models, a real-time multimedia streaming transmission control mechanism upon edge cloud computing and the opportunistic approximation optimization is proposed. Through the simulation and analysis experiments of static network topology and dynamic network topology, the performance of the execution cost, delay and packet loss rate of the proposed mechanism is deeply analyzed and verified. The analysis results show that the proposed mechanism can transparently influence the dynamic network topology and find an optimal solution for the guarantee of real-time and reliability through the deep fusion edge cloud computing and approximate optimization.