AI in computer networks
AI is no longer exclusively reserved for Hollywood movies and SF preachers. An exciting future awaits us, whether we like it or not.
While the title may be a bit pretentious, currently we live in what will be remembered as the beginning of an era in which machines can learn about themselves and their environment. The beginning of an era in which a human being does not have to be an ingredient of every recipe.
More and more SD-WAN networks (software-defined WAN networks) that use advanced algorithms for processing information are implemented with the goal – routers themselves making decisions without human factors. The aforementioned algorithms and related technologies are moving forward at a high speed. This speed is capable of disrupting the development plans of even the world’s largest IT companies.
Appropriate illustration of this hypothesis is the recent acquisition made by the world’s largest communication giant, Cisco Systems. They bought a do-it-yourself startup Viptela for $ 600 million. For Cisco this is a trifle, for the startup company is means big money. The fact speaks about the importance of this acquisition. When put into the technology context, it is possible that Viptela will slowly lead to the obsoleteness of Cisco’s till yesterday innovative solution for SD-WAN. The conclusion presents itself. The progress in the field of machines ‘ „self-learning” does not circumvent the computer network.
But let’s go back a bit and explain the idea behind the SD-WAN solution. Cisco has been trying for years to create a solution that will allow the user to really use backup links simultaneously with primary links. Additionally, the emphasis was on real-time performance monitoring and the ability of the network itself to respond to anomalies and problems by balancing traffic between links as needed. It must be acknowledged that they were the first to begin, but today there are many more versions of this technology – that of big players and of new companies such as Viptela.
The fundamental lack of Cisco’s version has been reflected throughout the complexity of the solution. In order to achieve the ultimate goal, the network administrator had to go through a lengthy router setting process, monitor regularly what is happening and possibly change some parameters as needed. It does, however, tackle the premise of the “Zero-Touch” solution, which stands in the subtitle of each SD-WAN document. Some authors cite this approach as “Minimal-Touch” and to be honest, until recently, none of the giants of the IT industry failed to do anything better.
In addition, it’s requiring certain series of routers to be used which in some situations may impede the greater use of this technology. If a user already has a set up WAN network with Cisco routers that works great but do not support this technology, replacing all of the equipment just because of the SD-WAN solution is not a real option.
With the development of cloud-based solutions, the zero-touch model is no longer so unattainable. For example, this has been proven already with Cisco Meraki implementations. Viptela tried and succeeded in implementing the concept on the WAN network. The goal was to reduce the time needed for the initial setup to a minimum and to move the control plane to the cloud. This way, the user is left to simply set the routers to the desired locations, turn them on so they can connect themselves to the controller which will continue to manage their work.
What does the final solution look like in the Viptela case? The control console and the interface, through which adjustments, analytics and controls are performed, are located in the cloud. This means they are always available from anywhere. Routers and controllers are deployed at user locations and are connect to each other without the engineer’s hardship in setting up everything. Apart from automatic initial setup, the main focus is of course on “self-learning” and adjustments of the network in real time without human engagement.
So, at the end of this text we are returned to the very beginning, to the title. AI is no longer exclusively reserved for Hollywood movies and SF preachers. An exciting future awaits us, whether we like it or not.