The project described focuses on analyzing the impact of the COVID-19 lockdown on the 800 MHz frequency layer of an LTE network in Milan. Here are the key points:
-
Clustering Analysis: The study began by clustering cells based on their traffic behavior throughout the week, using the Median Weekly Signature method. This involved analyzing data traffic volumes and categorizing cells into clusters that showed similar traffic patterns.
-
Cluster Identification: Three clusters were identified, with one resembling business area behavior (high traffic during working hours, lower on weekends), another reflecting residential area behavior (lower daytime traffic, higher in the evening), and a third cluster that was initially unclassified due to unexpected data patterns and was eventually neglected.
-
Geographical Mapping: The clusters were mapped geographically to visualize their distribution across Milan, with business clusters concentrated in the city center and residential areas spread around it.
-
Monthly Trends Analysis: Monthly traffic trends for each cluster were examined to observe changes over time, especially focusing on the period before and during the COVID-19 lockdowns in Milan.
-
COVID-19 Effects: The study detailed the effects of COVID-19 restrictions on network traffic, showing a decrease in traffic in business areas and an increase in residential areas during lockdown, as people stayed home.
-
Conclusion: The findings largely aligned with expectations based on class teachings and intuition about traffic patterns during the lockdown, despite some anomalies in cluster behavior that could not be fully explained.
This project was a joint effort with my colleague Davide Andreotti, we worked together on developing code, analyzing data, and generating visual aids to support our results.