The Design And Implementation Of A Simulation System Based On Virtualization Technology

Nowadays computer communication network plays an important role in industry and social life. For business needs, more and more big companies have built their own communication network. Power Company is the country’s basic industry, and it has a great responsibility to offer power for the country’s production and living. Now the company’s data communications network grid has developed to a large scale, and it faces enormous challenges of network maintenance, planning and upgrade. Network simulation system can provide a big help on network failure drills, network operating person training and so on.NS-2, OPNET are mainly used on network capacity analysis and network protocols tests. The simulation nodes of network do not have commands and physical characteristics of real network devices, so it is difficult to use these software on network failure drills and operating person training. Emulab network simulation platform has a higher degree of simulation. But in view of its implementation principle, when simulating a large-scale network it requires a lot of physical equipment. When enterprise build such a platform alone, it will take a high economic cost. For these reasons, we designed and implemented a network simulation system based on virtualization technology.Based on a detailed analysis of the functional requirements of the system, this paper completes the overall design of the network simulation system. The system is based on Dynamips and QEMU virtual simulation software. Its simulation node has the physical characteristics and commands of the real network devices. With a graphic user interface, the system is easy to operate. In order to support large-scale network simulation, the system uses a distributed system architecture. Combined with Dynamips, QEMU and other simulator, this system can use fewer physical network simulation equipment to complete large-scale simulation. This paper presents a mapping strategy based on topology graph partitioning for solving simulation network mapping problem. Besides, the paper also implements network traffic simulation and analysis, network fault simulation, network configuration and compliance analysis and other functions to help users analyze the structure and status of the network. After verification scenario, it proves that the system has good interaction and higher fidelity. It can be used to power communication network planning and research networks, network failure drill, network operation and maintenance person training and so on.