Sweeping the IPv6 Internet: High-Efficiency Router Interface Discovery With Weighted Sampling

Acquiring router interfaces is crucial for network measurement and security assessment. Established methods are readily available for IPv4 systems; however, efficiently pinpointing IPv6 router interfaces remains an unresolved issue, chiefly due to the vast IPv6 address space. Existing practices in this domain commonly suffer from inefficiencies. Thus, it is imperative to propose a methodology for fast enumeration of IPv6 router interfaces on a massive scale. In this study, we introduce Sweeper, a novel asynchronous IPv6 scanner that excels in discovering router interfaces from scratch, from few to many, on large-scale IPv6 networks. Unlike existing approaches, Sweeper requires merely the readily accessible IPv6 prefixes instead of seed addresses and can strategically optimize its probing direction based on a novel weighted sampling algorithm to increase the discovery rate. Real-world tests prove that Sweeper outperforms state-of-the-art works, discovering $33.2\%\sim 48.8\%$ more IPv6 router interface addresses than the baselines, with same computational resources. With Sweeper, we have collected approximately 6 million IPv6 router interface addresses from a single vantage point within less than one hour.