Empirical analysis of AS-level cooperation on the internet considering geopolitical characteristics

The Internet is a complex socio-technical system intricately embedded within physical geography and political contexts. While previous studies have emphasized the technical aspects of its topological structure, the geopolitical dimension of the Internet has not been fully explored yet. In this paper, we integrate multiple sources of data - including AS relationships and their national or regional affiliations - to study the Internet at the Autonomous Systems (AS) level, with a particular focus on cooperation patterns among ASes. Our findings reveal that the global AS network exhibits moderate disassortative mixing and weak nestedness, suggesting a flattening trend. However, within many individual countries or regions, AS cooperation patterns display stronger disassortativity and greater nestedness than the global average, indicating a more notable hierarchical organization. We further propose a visualization method to track the changes of cooperation trends among countries/regions. Finally, through case studies of the Russia-Ukraine conflict and Pelosi's Visit to China's Taiwan Region, we show that international AS-level cooperation (particularly for the countries/regions involved) may undergo rapid and significant changes in response to geopolitical events.