AI Used as the First Step in Automating IT

Print Friendly, PDF & Email

In this special guest feature, Huw Rees, Business Development at KodaCloud, discusses how applying AI to Wi-Fi is the first logical step to automating IT and indeed is the only way to realize the concept of optimization of Quality of Experience for every device connected to every access point in real time. Prior to joining the company, Huw held several senior leadership roles at 8×8, Inc., including vice president of business development, VP of Sales & Marketing and CEO of Centile, Inc., an subsidiary of 8×8. He received a B.Sc. (Hons) from the University of Manchester (England), Institute of Science and Technology in Electrical and Electronic Engineering and has an MBA from the University of LaVerne (USA).

IT automation is coming and by using AI, it can happen fast. If you think about it, businesses buy electricity, gas and water as utilities, they don’t have people on staff to make sure these utilities work on a day to day basis, they simply rely on the utility provider to provide the service as advertised and with 99.99%+ reliability, so why then should networking (LAN and WAN) be any different? The answer of course is that traditionally networks have been unreliable and have needed significant “tweaking” to extract the performance required by the end users and the only way to do this was to employ professional IT folks to ensure the value add from the actual business purpose was not adversely affected by networking issues.

So, what are the common networking issues and how can we solve them with automation? Gartner has reported that 64% of networking issues are actually Wi-Fi issues and when asked to fix the problem, IT staff spend 90% of their time simply trying to find the root cause. Often the fix is easy but identifying the root cause can be frustratingly difficult. These Wi-Fi issues then seem to be the low hanging fruit to solve by automation before moving on to other networking faults.

When analyzing Wi-Fi problems, they can be broadly categorized into two buckets, 1) users or devices cannot reliably connect to the Wi-Fi and 2) if they are connected they are not getting the performance they need. It is important to note that these items are dependent on the device, i.e. a video camera will need high bandwidth, but a Wi-Fi bar code scanner in a manufacturing process does not need high bandwidth but absolutely requires perfect connectivity. The key here is that there are “profiles” for different device types.

Given that we have now defined that different device types need monitoring for different parameters and that to be sure we are doing the absolute best to check this all the time, in real time, then we could conclude that we should hire a lot of tier two Wi-Fi IT support folks to look at every device connected to every access point simultaneously (and never take a coffee break or lunch). Clearly this is absurd and cannot be done, but if we substitute the IT staff for artificial  intelligence within a big data environment, we can start to see how this real time monitoring can be accomplished. In order for this to work, each access point needs to report back in real time to AI in the cloud the status of every device connected to it. Note that this does not mean any user data is being reported back, just the low level radio and network data, meaning there is no added security risk in this configuration. The AI can then monitor the status of each device, compare it to the “profile” for the device type and establish if the device is in a good or bad state. If bad it can change radio parameters like cell size, transmit power etc. to ensure the device has the best Quality of Experience (QoE).

In summary, applying AI to Wi-Fi is the first logical step to automating IT and indeed is the only way to realize the concept of optimization of Quality of Experience for every device connected to every access point in real time. IT staff can then be freed up to focus on the differentiated business issues their organization faces rather than running networks. In broader terms this becomes another step in Network as a Service (NaaS) resulting in networks being delivered as a utility in the not too distant future.

 

Sign up for the free insideBIGDATA newsletter.

 

Speak Your Mind

*