How SDN can help Enhance Edge Computing

By CIOReview | Tuesday, December 4, 2018

Edge computing is a distributed computation technique in which data computation is done at the device nodes as opposed to the traditional computation process where a centralized cloud environment is used for data computation. It’s a concept of completing data processing at the edge of a network. The edge can be any Internet of Things (IoT) device like mobile phones, laptops, sensors, actuators, home appliances, and many other devices. The idea behind edge computing is to push data analysis and artificial intelligence processes closer to the edge of the network to make the computation process fast and efficient. Edge computing also helps in cost savings by reducing the numbers of CPUs in the cloud. Edge computing helps to enhance the network and performance efficiency as the amount of data traversal is reduced by a fair margin.

Software-defined networking (SDN) is an architecture that helps to make the network agile and flexible. SDN allows enterprises to have better network control by enabling them to respond quickly to changing business requirements. It allows network engineers to redirect traffic from centralized consoles without interfering with the individual switches in the network. There are three layers of SDN network, namely:

Application layer: Application layer typically contains the network application and functions which include an intrusion detection system, load balancing, and firewalls. The application layer of SDN uses the application to manage data-plane behavior.

Control Layer: The control layer as the name suggests, acts as the brain of the SDN network. Centralized SDN controller software which resides on a server, manages policies and flow of the traffic throughout the network.

Infrastructure Layer: Infrastructure layer comprises all the physical switches in the network.

Software-defined networks help to combine the workforce of traditional clouds and edge computing. It acts as a deciding factor on whether the task is suitable for traditional clouds or edge computing. The Artificial intelligence enabled SDN controllers can identify the links which experience high network utilization on a specific period. The controller then sends information to process more data at the edge to eliminate network deadlocks.