Router-level Internet topology influences the dynamics of routing protocols, the scalability of multicast, and other aspects of protocol performance. In general, ISPs are reluctant to share their topology information with public due to commercial concerns. Although there are some studies to reveal underlying Internet topology maps, these are not very detailed. This work, to our knowledge, is the first automated topology data collection system built for accurately discovering the link level map of the Internet backbone. In this study, we designed and implemented a robust, modular, efficient, and dynamic system, named Cheleby, which collects Internet topology information. The system is deployed on PlanetLab platform and utilizes hundreds of systems scattered around the world. Cheleby is designed to meet the topological information needs of the researchers working on the Internet measurements, especially those who need link level topology maps of the Internet. In this thesis, we present the challenges in link level topology mapping, propose solutions to Internet topology data collection challenges, and explain our system which is designed to be robust, modular, efficient and dynamic.