Measuring Wi-Fi Performance from Mobile Devices

Measuring Wi-Fi Performance from Mobile Devices

Measuring Wi-Fi performance from a mobile device is easy, there are lots of mobile apps that can do this. Measuring is the easy part. The hard part is aggregating and analyzing that data on a network-wide basis.

Finger on the pulse of the Wi-Fi user experience

There are lots of apps for mobile devices ranging from disappointing to pretty cool, claiming to measure network or (mostly) internet speed. Some even keep a bit of history. For example there is Ookla’s app (> 1M downloads and ~4.5 stars), or there is the Wifi Analyzer and Surveyor from ManageEngine. These are handy, and free, and there’s a dozen more, just like them. But what are these apps really measuring, and what can they really tell you about the Wi-Fi experience for your users?

Wait, back up a second. Why should you want to measure the user’s Wi-Fi experience at all?

Well, when you know what Wi-Fi performance is truly like in every area of your premise or campus, even remote offices, you have powerful information.  You can plan network growth based on hard evidence about the Wi-Fi experience, not speculation. And you can proactively manage performance trouble spots before users are impacted. You’re in charge. You’re not a slave to the support line and the squeakiest wheel syndrome.

OK, so how do you obtain this information?

Performance is the missing ingredient in WLAN management

Measure WLAN performanceChances are, you may be considering Wi-Fi performance testing apps, because you’ve discovered you can’t get this type of information, in simple-to-interpret terms, from conventional WLAN management systems. And while spectrum analyzers can be invaluable for analyzing RF issues, they shed little light on the real-world Wi-Fi user experience on a humble smartphone in terms of performance metrics or a quality score.

So you try out a few tools and you find one that you like. Now what, how can you make use of it in an enterprise context? There are multiple requirements to consider: Frequency and Coverage, Reporting and Analytics, Data Relevance.

Frequency and Coverage

By “frequency” I mean how frequently you can collect the data not the radio band. Frequency and coverage are joined at the hip. You want to collect data from everywhere or a regular basis, right? Here’s the problem, if the performance testing app requires user intervention to run the test, then it is really only useful to IT for troubleshooting – perhaps by users guided by the help desk during a support call, or by an IT professional on his rounds.

What use is one isolated data point? If there’s a congestion problem or intermittent RF issue, you need to monitor performance over time. If the cause of the problem is device or driver related, one data-point won’t reveal that. You need data from multiple devices, sources and times to discover patterns and trends.

To be truly useful, the app must therefore gather data without user intervention – it must be automated to run tests periodically and build history. Once you have the ability to run tests automatically, it is hands free, and that means you can capture data regularly from wherever users happens to be. So this solves your requirement for network-wide data collection, too.

Reporting and Analytics

So far, all the apps we’ve found are stand-alone. A few have automation and history, but none have a supporting analytics framework for aggregating the data. So how do you get the data to IT for analysis? You could run scripts to parse the data from the test results on a web form, instead of using an app, or you could even write and app of your own. Or, some apps such as Ruckus’ SpeedFlex will package the results up and email them as a PDF to someone. Better, but still not enough.

A random snapshot is not much use except for troubleshooting, and most network administrators would be  reluctant to choke up their inbox with hundreds of snapshots from users, especially if the reports are for a healthy network. No, what you need is exception reporting, history and the ability drill down on-demand and examine the detail for any area of the network that looks dubious.

This requires a back-end with which to process loads of uploaded data, filter out the noise, and present only the interesting stuff in a meaningful form. The cloud can make this easier, giving you access from anywhere, and allowing easy scaling of the databases. But is the back-end development something you want to get into?

Data Relevance

Finally, there’s the matter of data relevance. What are you really measuring?  As businesses become more mobile, and move more of their mission critical apps to the cloud, maybe you should be monitoring the Wi-Fi experience for users connecting to those resources wherever they are, the cloud, your data center or even on the local LAN, not an irrelevant server in Seattle, or New York? Sure, a generic speed test tells you something, but how much of what is being measured is Wi-Fi performance versus Internet delay.

To distinguish between Wi-Fi, network and Internet you need to run tests against multiple servers in-network and in the cloud. You should be able to run tests against the servers which are mission critical to your mobile workers, not an arbitrary test server. To isolate network issues from Wi-Fi issues, you need over-the-wire control-tests and the ability to do internal testing only without touching the Internet.  You also need a combination of bi-directional throughput, round trip delay and voice quality tests, in order to get the full picture of the Wi-Fi user experience.

7signal’s approach to the problem

These are challenges 7signal has been solving since 2010. First we introduced Sapphire Eye fixed sensors which are perfect for monitoring the Wi-Fi performance for hundreds of clients simultaneously, in dense or business-critical areas. Next, we added the Micro Eye, enabling IT to monitor any area on an ad-hoc basis at short notice. Its Over-The-Air (OTA) reporting also makes the Micro Eye great in unwired areas too. Sapphire Cloud aggregates the data from the sensors and displays a dashboard of Key Performance Indicators representing the Wi-Fi performance SLAs you have set.

We had the inputs, we had the analytics, but still many areas of our customer’s networks remained unmonitored. For example, enterprises with lots of remote offices simply didn’t have the IT resources to deploy sensors unless they were visiting that location for another reason. Smaller sites often do not have the user density to justify a dedicated sensor.

For the complete picture, our customers needed a way to capture performance data from places sensors were not installed, as well – a testing agent that would go wherever users go, on the very devices whose Wi-Fi performance we want to track… Sapphire Mobile Eye was born.

Introducing Sapphire Mobile Eye

Sapphire Mobile Eye is the industry’s first enterprise-class Wi-Fi performance measurement app for iOS and Android devices. Used together with 7signal’s Sapphire Cloud Wi-Fi Performance Management service, enterprises now have the means to properly measure and visualize the Wi-Fi user experience in every corner of their networks, by exploiting all the smartphones, and tablets already in the workplace to collect Wi-Fi performance data everywhere users go.EyeQ Dashboard

The Sapphire Mobile Eye app will soon be here!  

Sapphire Mobile Eye turns BYOD to your advantage – In no time you could have a silent army of testers among your mobile workforce, automatically telling you exactly what the Wi-Fi experience is like, everywhere they go. Stay tuned for more information on Sapphire Mobile Eye as we get closer to our launch.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.