Global flow level network simulation method, system and device

Abstract

The invention discloses a global flow level network simulation method, system and device, and the method comprises the steps: constructing a global simulation network topology architecture to form a simulation environment, generating a simulation task through a task generator, dividing the simulation task into a plurality of stages, and storing each stage of the simulation task into a global variable for global call; the method comprises the following steps: continuously updating time information transmitted by flow objects in a global simulation network topology architecture through a global refreshing strategy, determining a flow event of which the time is shortest for each global flow object, and after the current flow object completes the flow event, calculating flow event completion time expected values of other flow objects through refreshing; global stage information is obtained through a global control strategy, whether all flow objects in a certain stage complete flow events or not is detected, so that the end of the stage of the simulation task is predicted, and a test result of the simulation task is obtained. According to the network simulation scheme provided by the invention, the resource consumption is reduced, the calculation time is shortened, and meanwhile, the simulation correctness is ensured.

Description

technical field [0001] The invention relates to the field of simulation technology, in particular to a global flow level network simulation method, system and device. Background technique [0002] The research and development of data center network, World Wide Web communication, and wireless communication technology largely rely on simulation technology, including link simulation, system simulation and network simulation. The simulation scale of network simulation is relatively large, usually tens of thousands of nodes simulation. The existing network architecture is mainly servers, switches, and links. In order to cope with the increasing traffic and increasingly complex application requirements in the future, it is often necessary to conduct simulation verification analysis before deploying or upgrading the real network architecture. . [0003] The inventors of the present application found that the existing network simulation technology has the following defects: [00...

AI Technical Summary

Problems solved by technology

[0004] (1) Discrete event network simulation at the packet level, represented by NS3 and OMNeT++, is suitable for fine-grained simulation of hierarchical protocols. However, due to the dependence on the packet level and event-driven framework, when performing large-scale multi-node multi-flow simulation, Consumes too many resources and time, even if the distributed simulation function is turned on, it will take days or even months

Based on this, machine learning is derived to replace some simulations of similar network structures. For example, MimicNet (Sigcomm 2021) only allows OMNeT++ to simulate the network structure of a specified unit, and put the network structure of the remaining units into the Mimic approximation module. calculation, so as to achieve a certain degree of speed improvement, but this method cannot accurately approximate the singularity of the network structure and the burstiness of the traffic

Currently, some methods simplify a part of the protocol stack, such as Netbench, but are still limited by the package level and event-driven framework

[0005] (2) Using a kernel-isolated virtualized network emulator, taking Mininet, IM∪NES, and CORE as examples, is only suitable for small-scale simulations that can control each node in real time, but a single machine has limited resources and cannot simulate a large-scale network

[0006] (3) Using a simulator combined with a real switch, such as TurboNet, which can simulate a switch into an 8-node fat tree structure, but it still cannot do large-scale simulation in the true sense

Images