A lot of the time last week was spent on writing the design report. I started working on getting the front end of the project working, which is to do packet sniffing to identify access points, send beacon frames with TIM set up such that all devices connected to that access point responds (in order to quickly get a list of devices in the vicinity), and using those devices’ MAC address to get data such as ToF or RSS. Wi-Peep’s paper suggests that exploiting the TIM should get responses from the devices within the access points fairly quickly, but it seems that my setup is somehow incorrect since it doesn’t reliably work. I will work on getting these parts done next week, along with using the devices’ MAC addresses and picoscenes to get RSS data.
Anish’s Status Report for 03/11
A significant chunk of time in the past week went into writing up the design review report and documenting our plans for the project. Because of limitations on timestamp capturing in PicoScenes that I don’t think we’ll be able to work around, I started looking into how we can use an ESP32 microcontroller to get similarly-precise timestamps, like is done in the Wi-Peep paper. I wrote some code to run the ESP32 in monitor-mode and take timestamps when the interrupt is received. I took some test measurements and noted it was significantly noisier than the Wi-Peep paper’s results; it also seems to vary quite a bit with changes to the optimization and driver settings for the ESP. Next week I will be looking further into if there are any more build-time settings that can be changed or other ways to improve the measurement quality.
