WiFi analytics is exploding in use as everyone realizes just how crucial data analysis is for understanding and enhancing network capabilities. This technology provides valuable insight into network usage and performance, simplifying WiFi upkeep and network optimization. The features and technology available differ from company to company, with solutions using a combination of hardware, software and Cloud-based services. Here are a few things to consider when getting started with WiFi analytics.
Wired, Wireless, or Both
Analytics are only as valuable as the data being analyzed. The more information the better – as long as the right information is singled out and targeted, something discussed in ‘The Cloud’ section below. Broadly speaking, WiFi analytics works by tracking wireless/wired devices connected to the network, devices nearby the network, and/or network hardware, such as access points. Deciding which analytics to use depends on what purpose the analytics will serve.
Using sensors or APs to collect data about non-connected clients works for companies that want metrics on device type or number of unique visitors, but don’t need detailed information on user experience.
Collecting data from connected wired and wireless devices increases insight into the network, and provides a clear view of what clients are experiencing. However, it doesn’t provide complete network visibility.
Analytics from all infrastructure equipment, wired/WiFi-connected devices, and non-WiFi devices (e.g. Bluetooth) provides complete network visibility.
Some companies use unique identifiers, or fingerprints, for tracking devices. These fingerprints can identify hardware types, operating systems and software versions. With this information, companies have a deeper understanding of the devices on their network and can understand if changes on the network impact one group of users more than another. This type of analysis is only available when using device fingerprinting.
Crawlers and Radios
Now into the details: how is the information captured? Once again, there are a few different methods currently on the market.
Many search engines use crawlers to gather content from publicly available websites. In WiFi analytics, the concept is similar, although the application is slightly different. With a software crawler, analytics companies intercept network data through a span, tap or monitor port on a network switch. The captured packets are then analyzed for relevant data.
Other platforms use access points with multiple radios for data capture. The radios collect data, perform analysis, and run tests. Here are some things to look for to differentiate radio technology between companies:
Vendor-agnostic: Leading edge sensor technology should be vendor-agnostic so that it works with any and all hardware and software, now and in the future.
Manual or Automatic System Tests: Several companies offer the capability to run network tests. During this scenario, a radio has the ability to act like a client, connect to the network, and run several different kinds of tests. One of the most common options is a speed test. Companies who configure their radios to run these tests automatically take this capability to the next level. Automatic tests free the IT department from traveling to remote sites without losing any data insights.
Full Spectrum Analysis: Look for radios with the ability to capture and analyze WiFi and non-WiFi data.
Automatic Notifications: Automatic data capture and analysis is further enhanced by a system that can perform automatic self-analysis as to why any tests failed, and then automatically notify system administrators with the analyzed problem and actionable steps for resolution.
The Cloud, Machine Learning and Big Data
Companies use the Cloud to provide scalable, instantly shareable information analyzed with the use of machine learning and algorithms. Be sure that any Cloud infrastructure only receives pertinent data. Data capturing tools can see billions of packets a day. Sending all that data to the Cloud for analysis only increases network overhead. Look for companies using “edge computing” or “fog computing.” These principles outline how to analyze data before it reaches the cloud. This leaves the Cloud infrastructure free to run deeper analysis and combine data from multiple sources and locations – providing rich analytics.
Machine learning and big data algorithms are the secret to WiFi analytics success. These capabilities enable software to answer complex questions and provide complete network visibility. They move platforms beyond single aspect analysis and provide detailed understanding of a network as a whole. Look for companies that use machine learning algorithms to support real-time updates and proactive system changes. With this capability, WiFi analytics should result in detailed analysis, automatic testing and reports, and proactive alerts. Remember, WiFi analytics shouldn’t simply report a problem – it should provide a solution as well.
Wyebot uses the market’s current best 802.11ac Wave 2 radio, providing unparalleled range and performance. We solve current problems with automatic, real-time analysis and future problems with the proactive notification of changing capabilities and interoperability. We provide business analytics for capacity planning with historical data trends. Our vendor-agnostic platform is the only one on the market to provide full packet capture and automatic analysis using an Artificial Intelligence engine. We are the leaders in WiFi Assurance. Contact us for a demo today.