WLANMON – Measuring Wi-Fi from the Client’s Perspective: The Idea Behind It

“The Wi-Fi is slow again.” Anyone who looks after wireless infrastructure knows that sentence. It usually arrives without a time, a location or a device. And when you check, everything is green. In this article I describe the idea behind WLANMON: a small measurement system that continuously observes the Wi-Fi from a client’s point of view – and why I believe this is exactly the perspective that is often missing in day-to-day operations.

The Problem: Two Views of the Same Network

Modern WLAN solutions come with extensive management systems. Controllers, clusters and network management systems show which access points are online, how many clients are connected, how busy the channels are and whether there is interference. That is valuable – but it is the view of the infrastructure.

Users experience the Wi-Fi from the other side: Does my device see the SSID? How long does it take to connect? Do I get an IP address? Does DNS resolve? Can I reach the gateway and the internet? An access point can be formally “up” while authentication is stuck at the RADIUS server, the DHCP pool is exhausted, or a client in a certain part of the building simply has too little signal.

Most discussions happen right in the gap between these two views.

Snapshots Are Not Enough

The classic troubleshooting tools are site surveys, a laptop running analysis software, or a technician on site. They deliver good data – but only for the moment someone is measuring. Many Wi-Fi problems, however, are intermittent: at 8 a.m. when everyone arrives at once, after a firmware update, only in one particular room, or only when the neighbour puts a new access point on the same channel.

And of course, as soon as the technician shows up, everything works fine.

The Idea: A Sensor That Behaves Like a Client

The basic idea of WLANMON is simple: a small, inexpensive device is permanently placed at a relevant location and behaves just like an ordinary Wi-Fi client – except that it logs every result.

Periodically, the sensor essentially does two things:

• Scan the environment: Which SSIDs and access points are visible, on which channels and bands, at what signal strength? This shows how the RF environment changes over days and weeks – including neighbouring networks.
• Test the connection: The sensor actively connects to defined SSIDs and checks the same chain a user goes through: association, authentication, DHCP, DNS, reachability of the gateway and targets. Every step gets a timestamp and a duration.

Many such individual measurements add up to a history. “The Wi-Fi is slow” becomes “since Tuesday, the DHCP step on the second floor regularly takes more than five seconds.”

Why Centralised?

A single sensor is useful, but things get interesting with multiple measurement points – in different parts of a building or at different sites. That is why WLANMON is designed in two parts from the start:

• The sensors measure on site and send their results to a central server.
• The server collects the data, presents it in a web interface and manages the sensors’ configuration – which SSIDs to test, at what interval, against which targets.

This way nobody has to touch each sensor individually, and results from several sites can be compared side by side.

Deliberately Simple Hardware

A sensor does not have to be expensive. A small single-board computer running Linux and a good USB Wi-Fi adapter are perfectly sufficient to start with. That keeps the cost per measurement point low – and with it the barrier to measuring in several places instead of just one. The WLAN Pi community provides an excellent foundation for this.

Commercial sensor and “synthetic monitoring” solutions do exist, but they are often tied to a single vendor or simply oversized for smaller environments. WLANMON is meant to be vendor-independent: the sensor simply measures what a client experiences – no matter whose access points are mounted on the ceiling.

What Is It Good For?

• Evidence instead of gut feeling: Complaints can be backed up with data – or refuted.
• Baseline: Knowing how the Wi-Fi behaves under normal conditions makes deviations easier to spot.
• Before/after: Firmware updates, channel planning or new access points can be assessed objectively.
• Early warning: A sensor often notices a broken SSID before the first ticket comes in.
• Customer conversations: A history covering several weeks is a far better basis for discussion than a single measurement.

Conclusion

WLANMON replaces neither a management system nor a professional site survey. It complements them with the perspective that ultimately matters: the client’s. An inexpensive sensor that honestly records, around the clock, what an end device experiences answers many questions before they turn into problems.

The project is available on GitHub: github.com/dgawin/wlanmon.

In future articles I will go into the technical design, the hardware and the analysis in more detail.

Note: WLANMON is a personal project and is not affiliated with my employer SPIE.