A Raspberry Pi with Wifi & Bluetooth Probe Feature
Jonas Escobar and Yixiu Zhu
Explore the docs »
Table of Contents
We decided to offer our Raspberry Pi with a wifi and bluetooth probe feature: it can collect WiFi hotspots' and bluetooth devices' communication requests within a specfic range. Note: all code was provided by Professor Michael Hersch, PhD, from Boston University. We were responsible for using his code on the Raspberry Pi to get some understanding of the Pi's functionality.
Here are our intended features:
- detect Bluetooth legacy devices in the area with its device name
- detect BLE devices in the area with its device name
- detect wifi hotspots over time and put into a JSON file
- plot wifi hotspots over time ("Passenger flow volume detection")
Addressing why choose raspberry pi and its performance & attachments:
Raspberry Pi is a powerful embeded system for educational researches and Pi 4's design offers capability to connect to monitors for better visualization. Raspberry Pi OS is Python supported and it helps us to be familar to Python and its basic packages.
One of the main difficulties we had when setting up the Raspberry Pi was getting the correct setup. As an example, we tried to use a macbook as our monitor, but soon found that would not work, so we had to adapt and get a monitor with an HDMI cable to interact with the UI for the Raspberry Pi. Additionally, after setting up the SSH with the wifi on an on campus apartment. We had difficulties using the SSH from one of the BU dorms, namely because the wifi router at the apartment did not allow the computer in the dorm to connect.
In addition, we learned a lot about the syntax for the Raspberry Pi and Python 3 specifically. There were many times the code was not working because we were ussing the incorrect syntax. When running wifi_plot.py, the file was not working because an argument had to be called for the filename when executing the script.
python3 wifi_plot.py wifi_2021-09-16T12_02_16.jsonAlthough not many modifications were made because the code was given, we did make 2 small changes in order to increase functionality and get a feel for using Raspberry Pi.
First, we added to the bluetooth.discover_devices command in bt_scan.py, so that the script also returned the device name in addition to the bluetooth address.
disc_devs = bluetooth.discover_devices(duration=5, lookup_names=True) This was done primarily to see what devices we were detecting to get a feel for how the script operated. Namely, how the script only picked up on our phones and laptops when they were actively looking to connect to a bluetooth device.
The second addition was to wifi_plot.py. We added more labels to the plot, so it looked more professional and easy to understand.
matplotlib.pyplot.xlabel("Time (Day of the Month/Hour/Minute)")
matplotlib.pyplot.ylabel("Number of Hotspots")
matplotlib.pyplot.title("Hotspots over time in Photonics")
matplotlib.pyplot.show()
The Figure plots the hotspot traffic near the highway for 10 minutes. Data was collected in PHO 111 near the highway, which means there were likely some stationary devices in photonics that interfered with the traffic data. As seen in the figure, early in the data collection, there was more traffic, while traffic dissipated near the end. The code could be updated in order to separate out wifi hotspots that are stationary and those that are moving (ie. the cars moving in the highway).