SDN-SPS: Semi-Physical Simulation for Software-Defined Networks

Abstract

Nowadays, software-defined networking (SDN) is one of the key technologies promoting the development of the next generation network. In addition, simulating and building the SDN real network is one of the most important approaches to do SDN research. However, there are many aspects that limit the test platform-based experiment and the traditional software-only simulation, such as the various demands of network, complex applications, and flexible network architecture. Combining test platform with the software-only simulation, this paper proposes a new hardware-in-the-loop emulation: to emulate a SDN. We use the system-in-the-loop technology of OPNET to build the SDN hardware-in-the-loop emulation platform. The emulation test platform includes the sensor network part, all of which will use the software simulation, together with the transmission network and the cloud data center both of which will use the real network. We run the simulation to generate a few applications in the cloud data center, and then collect the data through the transmission network and the sensor network, which makes these applications share the sensing data. This paper proves that the simulation test platform of hardware-in-the-loop can achieve energy efficiency under the demand of QoS.