HotSpots

• 

  slides presentation

Inspiration

In recent months, students across the BU campus have experienced inconsistent Wi-Fi performance, especially during peak class hours. These network drops often occurred without clear visibility into which access points were overloaded or failing. We wanted to design a practical tool that could help the university proactively identify and respond to these weak spots.

What it does

We developed a Python-based data pipeline that collects and processes Wi-Fi association data to:

Map and visualize popular access points across campus

Detect usage intensity over time

Highlight areas that might need network upgrades or load balancing

We used datasets from multiple sources (including a historical dataset from Dartmouth) to build a prototype that simulates how this system would work in real time at BU.

How we built it

Ingest raw Wi-Fi connection logs (e.g., number of users, connection duration, floor data)

Original Dartmouth Data: https://ieee-dataport.org/open-access/crawdad-dartmouthcampus-v-2009-09-09

Original HKUST Library Data: https://dataspace.hkust.edu.hk/dataset.xhtml%3Bjsessionid%3D4bfaff0d7879eb6c127a020b4eec?fileAccess=&fileSortField=date&fileTypeGroupFacet=&persistentId=doi%3A10.14711%2Fdataset%2FRQXRQB&q=&tagPresort=false&version=&utm_source=chatgpt.com

Convert access point information into geospatial coordinates (real or synthetic)

Generate 2D and 3D heatmaps showing usage density

Flag high-traffic or overloaded APs for potential network action

Challenges we ran into

Handling incomplete or missing location data

Designing a placement algorithm that works with both GPS and floor-only datasets

Ensuring visualizations remained accurate and lightweight

What we learned

This project showed how data engineering and visualization can turn abstract network logs into actionable insights. With more complete live data, this system could help campus IT teams monitor and optimize Wi-Fi in real time.

What's next for Hotspots

With these two starting maps, HotSpots can generate heatmaps of network usage in any area (2D heatmap) or any building (3D heatmap) using the provided Wi-Fi data. Once HotSpots becomes accessible for institutions to upload their own Wi-Fi data, students and professors can make better decisions about their study or teaching spaces, and technicians can identify areas with high network traffic to improve data for their institution.

Built With

• dartmouth-wifi-data
• hong-kong-university-wifi-data
• python