Research On System Architecture And Service Placement Strategy Of Communication And Computing Cooperative Edge Computing

With the continuous innovation of wireless communication technology,the Internet of Things(Io T)has also begun to develop rapidly,and various types of Io T terminal devices are ubiquitous in our lives,providing great convenience for our lives.However,with the surge in the number and types of Io T devices,the high-frequency message data reported by devices has created a significant bandwidth burden on the cloud.As the Io T continues to develop,people have higher requirements for the service quality of Io T.However,the transmission delay for Io T devices to obtain computing services under the cloud computing mode is large,and once data communication with the cloud is lost,the device will lose all computing capabilities,causing the entire Io T system to become paralyzed.To address these issues,using edge computing architecture as an alternative to cloud computing architecture has become the optimal solution.However,the current applications of edge computing are suitable for specific scenarios and have not provided a universal solution for Io T scenarios.This thesis aims to propose a universal communication and computation fused edge computing architecture for various Io T scenarios,and to study a joint resource allocation service placement algorithm for Io T scenarios after introducing edge computing.The contributions of this thesis mainly include the following aspects.1.A universal communication and computation collaboration edge computing architecture for heterogeneous Io T systems is proposed.The edge computing platform under this architecture is designed with a microservices approach,allowing decoupling between edge computing platforms,and inter-module communication through message buses.The interaction mode between internal modules of the edge computing platform is also designed,enabling dynamic expansion of the edge computing platform,and deployment of corresponding computing services according to scene needs to achieve specific functions.In addition,container and container orchestration technologies are used to deploy services,and the combination of container technology and decoupled edge platform enables modules to be plug-and-play,allowing dynamic adjustment of service modules of the edge platform.The edge computing platform can also simultaneously interface with multiple Io T systems,capable of handling the access of heterogeneous Io T devices.The edge computing platform can import device attribute update information into the computing module,and the computing module can also complete device control,realizing the fusion of computation and communication.2.A service placement algorithm is designed for Io T edge computing scenarios with limited resources.The algorithm considers the communication,storage,and computing resource limitations of edge computing servers,and combines the resource allocation of edge computing servers to optimize system energy consumption while meeting user service quality requirements.The optimal solution for service placement and resource allocation decisions is obtained through mutual iteration using convex optimization and genetic algorithms.Simulation analysis shows that the algorithm can meet user service quality requirements and has lower energy consumption than traditional algorithms.3.A prototype system is built to verify the feasibility of the proposed edge computing architecture,and test scenarios are designed to test system performance.Testing shows that the architecture can address the heterogeneity of Io T,dynamic service deployment,and improve user service quality,providing reliable computing support for different Io T scenarios.