Performance Benchmarking of Software Defined Networks: A Study of Ryu Controller with D-ITG

Software Defined Networking (SDN) is a novel paradigm in the networking industry that is implementing redefined goals of traditional networks. With SDN, inflexibility and complex networks will be replaced by a new flexible and programmable network. In SDN, the data plane and control plane are kept apart and the traditional network constraints are reduced. The SDN controllers like NOX, POX, ONOS, OpenDaylight, Floodlight, RYU etc. are acting as the network brains and oversee its operations. The forwarding decisions are sent by the controller to switches and routers. The forwarding of the packets is done by switches. Network traffic generators are widely used in networking research to verify results and evaluate the impact of novel protocols and network features and algorithms. In this research paper, Mininet is used for SDN emulation. The performance is evaluated for different types of network traffic. D-ITG (Distributed Internet Traffic Generator) is used for traffic generation. A custom topology is created in Mininet using Python. The network performance metrics (bandwidth, throughput, round-trip time, jitter, and packet loss) are evaluated for a variety of traffic types, including TCP, UDP and VoIP.