Research On Edge Computing Based On Mobile Self-Organizing Network

Aiming at the ultimate demand for information acquisition in the depth,breadth and intelligence of the future,the intelligent sensing technology for large-scale dynamic ad hoc networks and collaborative computing is the key technology to break the microsystem aware network information intelligence.In this paper,the edge computing based on mobile self-organizing network is taken as the research topic,focusing on large-scale dynamic instant networking technology and dynamic programming problem based on edge computing model.The main research content is divided into three parts.Firstly,research and build the platform OPNET for large-scale dynamic real-time networking simulation,and use OPNET to study the core issues such as packet format,neighbor detection and MPR node set of mobile ad hoc network routing protocol OLSR and improve OLSR:(1)The serial number of the new and old routing information can be distinguished,and the end-to-end delay is reduced.(2)The multipoint relay(MPR)mechanism is adopted to shorten the length of the control packet.Then,the simulation verifies and optimizes the large-scale dynamic instant networking success rate and network connectivity and network bandwidth and delay,so that each node can forward the accusation commands to each other,exchange the sensing situation,health status and intelligence gathering data,and automatically connect to establish a network.Secondly,using the simulated mobile ad hoc network,the ultra-low communication delay and computation delay are realized by direct short-range communication across multiple edge computing nodes,and the low-latency design is transformed into a dynamic programming problem.In the trade-off between communication and calculation across multiple edge nodes,the joint scheduling of communication and computation is transformed into the double-catch problem.A cooperative task calculation algorithm driven by various delays is proposed,so that when edge nodes are selected at the same time,Provide appropriate heterogeneous resource allocation for multi-user services.Finally,combined with the system simulation tool of Edgecloudsim edge calculation,using the software proposed workflow,the sensor and camera analyze the system processing speed,task failure rate and network delay in the situation-aware scene.In the mode,comparing the results of data processing in the three modes of cloud,edge,cloud and edge coordination,the advantages of edge computing are summarized,and the system performance analysis under two different collaborative modes reflects the superiority of the algorithm.