Statistics Measurement of Network Traffic in Software Defined Networking

Abstract

The recent developments in existing network infrastructure need high bandwidth capacity and multiple paths between high-end switches. Software-Defined Networking (SDN) provides QoS for network flows. Statistics measurement is an essential task in traffic engineering. Traditional equal cost multi-path (ECMP) routing used to statically steer flows among multiple equal-cost paths but it cannot guarantee optimal resource utilization. This static hash collision can degrade throughput and high latency for the flows. Flow Path Computing algorithm (FPCA) can minimize the network congestion by rerouting the flows over the alternative paths in SDN. The FPCA algorithm evaluates the existing flow's demand based on port statistics and the existing flow is shifted from the congested path to the alternative light loaded path. In this paper, this paper also takes a look at statistics measurement in SDN to improve the performance of HTTP and FTP traffic over ECMP. The evaluations indicate that FPCA algorithm enforces the required end to-end QoS for each traffic flow.