IP geolocation is essential for the territorial analysis of sensitive network entities, location-based services (LBS) and network fraud detection. It has important theoretical significance and application value. Measurement-based IP geolocation is a hot research topic. However, the existing IP geolocation algorithms cannot effectively utilize the distance characteristics of the delay, and the nodes' connection relation, resulting in high geolocation error. It is challenging to obtain the mapping between delay, nodes' connection relation, and geographical location.
To solve the problems, a research team led by Fenlin LIU published their new research on 15 December 2024 in Frontiers of Computer Science co-published by Higher Education Press and Springer Nature.
The team proposed propose a representation learning model for IP nodes (IP2vec for short) and apply it to street-level IP geolocation. IP2vec model vectorizes nodes according to the connection relation and delay between nodes so that the IP vectors can reflect the distance and topological proximity between IP nodes. Compared with the existing algorithms such as Hop-Hot, IP-geolocater and SLG, the mean geolocation error of the proposed algorithm is reduced by 33%, 39% and 51%, respectively.
In the research, they propose a street-level IP geolocation algorithm based on IP2vec. They use a neural network to fit the mapping relation between vectors and the geographical locations of landmarks to achieve geolocation.
Firstly, they measure landmarks and target IP to obtain delay and path information to construct the network topology. Secondly, they use the IP2vec model to obtain the IP vectors from the network topology. Thirdly, they train a neural network to fit the mapping relation between vectors and locations of landmarks. Finally, the vector of target IP is fed into the neural network to obtain the geographical location of target IP.
In addition, the IP2vec model can also be applied to other fields, such as the user geolocation of social software, the following relation between known users, and the distance between users displayed on social software. Input the above information into the IP2vec model, and infer the geographical location of other users according to some users with known geographical locations. Such research can be carried out in the future.
DOI: 10.1007/s11704-023-2616-9