Method and system for hierarchical distributed deployment of virtual network based on reinforcement learning

Abstract

The invention discloses a virtual network hierarchical distributed deployment method and system based on reinforcement learning. The invention is oriented to network simulation based on Docker containerized virtual network, and aims at virtual network optimization segmentation and deployment problems, in computing, network, storage In the distributed environment of physical hosts with limited resources, through the reinforcement learning framework, a reward mechanism that takes into account the resource consumption of physical hosts and the performance loss of cross-host communication between virtual network blocks is designed. The long-term benefits of taking different actions in different states enable the algorithm to dynamically and autonomously learn the optimized virtual network block size and appropriate deployment timing, and introduce a certain amount of dynamic randomness through the randomness strategy to avoid excessively rigid selection of the largest actions, resulting in the problem of excessive resource consumption or fragmentation, so as to achieve the purpose of virtual network layering and distributed optimization deployment.

Description

technical field [0001] The present invention relates to the technical field of network virtualization, in particular to a method and system for hierarchically distributed virtual network deployment based on reinforcement learning. Background technique [0002] Network simulation is the key support for computer network architecture, protocol, and algorithm research. Because the Docker container retains the basically complete TCP / IP protocol stack, and compared to the virtual machine, the container has higher startup efficiency and less performance overhead, so the containerization technology is gradually becoming popular, and it also provides network simulation simulation A new idea—construct virtual network elements (such as virtual routers, virtual switches, and virtual end systems) with Docker containers as the core, and use technologies such as veth-pair, OVS (Open vSwitch), and VxLAN (Virtual eXtensible Local Area Network) Generate a virtual link for connection, form a ...

AI Technical Summary

Problems solved by technology

The present invention is oriented to network simulation based on Docker containerized virtual network, aiming at the problem of virtual network optimization and deployment, in the distributed environment of physical host with limited computing, network and storage resources, through the reinforcement learning framework, the design takes into account the physical host Reward mechanism for host resource consumption and cross-host communication performance loss for virtual network slicing ,according to Continuously calculate that each physical host takes different actions in different resource supply states The long-term benefits of , so that the algorithm according to Dynamically and autonomously continuously learn the optimized virtual network block size and appropriate deployment timing, and introduce a certain amount of dynamic randomness through the randomness strategy to avoid excessively rigid selection The biggest action will cause the problem of excessive resource consumption or fragmentation, so as to achieve the purpose of virtual network hierarchical, distributed and optimized deployment

Images