The model needs to generate two files: topology data representing AS-level connectivity, and IP prefix data used for longest prefix matching.

• Model 1: use the largest N ASes in the DIMES dataset.
• Model 2: use the most central N ASes from DIMES dataset.
• Model 3: extract ASes from different geographical areas from DIMES dataset.
• Model 4: based on jellyfish model and DIMES dataset. use jellyfish model to analyze the DIMES dataset, get the AS statistical distribution and zoom in the scale to smaller size that we can evaluate in proportion to the analytical statistics.