With the growing scale and complexity of the Internet, there is an increased need to understand and manage different aspects of this critical infrastructure. In particular, identifying and characterizing features of network traffic across the IP address space could provide valuable insights for managing the network and detecting network attacks. To this end, Hilbert Curves offer a powerful method for visualizing features of network traffic across the IP address space. However, prior studies that follow this approach have various limitations such as exploring a limited set of traffic features, considering limited datasets, and focusing only on data with a specific granularity.

In this thesis, we explore how visualization with Hilbert Curves can be used to examine traffic features and their correlation with the structure of IP addresses. To this end, we explore how Hilbert Curves can be used to properly visualize network traffic features at different granularity and to examine any correlations between the structure of observed IP addresses and the associated traffic features. We apply this visualization approach to three different real-world questions with large traffic datasets to gain insight into the structure of network traffic and the efficacy of using visualization. The developed software used within this thesis for visualization is publicly available to facilitate other researchers to examine and reproduce our analysis.