Hi,
I wanted to share a small update on WTF — Wi-Fi Test Framework.
After the first early version, I continued extending the tool from a basic throughput automation script into a more complete Wi-Fi/network testing CLI.
The current development version now supports:
-
cold RTT measurements;
-
loaded RTT measurements;
-
iperf3 throughput measurements;
-
CPU load measurements on both sides;
-
basic result plotting.
The main idea is to measure not only raw throughput, but also how the network behaves under load.
For example, the tool can now run a throughput test and measure RTT/latency behavior during that load, while also collecting CPU load from both the client side and the OpenWrt AP side. The goal is to make it easier to see whether a test result is limited by Wi-Fi conditions, CPU load, AP behavior, or something else.
The current target setup is still:
-
Linux client machine running WTF;
-
OpenWrt-based AP controlled over SSH/UCI;
-
iperf3 installed on both sides;
-
separate control and test networks;
-
configuration file describing AP/client addresses and test settings.
I’m planning to run more end-to-end tests on different hardware, including Wi-Fi 5 and Wi-Fi 6 OpenWrt APs, before moving toward the next larger version.
At this stage, I would especially appreciate feedback on:
-
whether the measurement workflow makes sense for OpenWrt-based Wi-Fi testing;
-
better ways to measure loaded RTT during iperf3 traffic;
-
possible OpenWrt-specific edge cases when changing radio/channel/bandwidth settings;
-
whether collecting CPU load from both sides is enough, or if other system metrics would be useful;
-
how to make the test methodology more reliable and reproducible.
The project is still early-stage, so I’m not presenting it as a finished benchmarking tool. I’m mostly trying to build a practical testing workflow and would appreciate criticism from people with more OpenWrt/Wi-Fi testing experience.
GitHub:
Any technical feedback is welcome.