A survey of scheduling policies in software defined networks
Anuradha Banerjee, Flavio Esposito · 2017
In the era of Software Defined Networks (SDNs), multi-policy resource management is extensively used to deliver ready-to-use media-optimized applications. Switches, ports and other shared resources are dynamically allocated to different flows, mostly based on priority. Such priority may be either externally set or computed depending upon various factors, such as flow package size, importance set by the user, age in the queue, etc. This resource schedule facilitates high speed communication under large scale distribution, efficiently manages the network bandwidth, and makes the resources available on demand while ensuring their efficient utilization. Keeping in mind the heterogeneity of network resources e.g., differences in capacity of handling workload, cost, energy consumption, and the (mostly) exponential distribution of flows granularity, a significant number of scheduling strategies has evolved in the literature of SDN. In this article we discuss such scheduling strategies, along with their advantages and disadvantages and we give a few research directions on the topic. To the best of our knowledge, this is the first article that focuses on scheduling strategies in SDN.