When you suspect an unknown device has latched onto your WiFi, the usual first step is to log into your router and check the list of detected MAC addresses. But there's a more advanced option that requires a bit of code and several Raspberry Pi units. The result is FIND-LF, a WiFi detection platform that physically locates devices within its range—even when they aren't connected to your network.
The Problem of WiFi Freeloaders
Few things are worse than finding someone mooching off your WiFi. In an ideal world, internet access would be cheaper, but providers keep hiking prices, pushing more users to steal connectivity, especially on mobile devices. A strong password and the best security protocol your router offers will discourage most intruders, but if suspicions remain, the router's list of MAC addresses provides a solid clue. The alternative is a far more elaborate detection method.
How FIND-LF Works
FIND-LF, created by Zack Scholl, builds a specialized network of Raspberry Pi receivers that constantly send 'probe requests' to every WiFi device within their reach. These packets are forwarded to a server that processes the devices' 'fingerprints' and then passes them to another server that uses machine learning to compute their physical location. The clever part: devices don't need to be connected to a local wireless network to be detected.
The official GitHub repository contains all the necessary details, scripts, and commands to set up the system, but we have to admit it's not a simple task. Tracking effectiveness depends on the number of receivers, so you'll typically need at least three or four Raspberry Pi units, each with its own power supply, a WiFi module configured for monitor mode, and internet access. Then there's the initialization of the Pis, which can be simplified with the help of PiBakery. In other words, tracking WiFi devices with FIND-LF is definitely an advanced project, but no less interesting for that.
