Network Measurement and Analysis
It is essential to network operations and technology development
to understand the usage and behavior of the Internet.
However, we can observe only limited part of the Internet and collected data is often huge but noisy. Thus, it is necessary to extract meaningful information from big data or to infer general trends from partial observations.
IIJ Research Laboratory is conducting unique analytic research
using various data gathered by collaboration with other organizations
as well as real data obtained from IIJ's own networks and services.
Our research includes traffic trend analysis, quality of service inference, and anomaly detection, by applying statistical analysis and machine learning techniques to a large volume of network data.
Our research results are widely recognized as valuable sources for understanding the Internet usage and behavior. For IIJ, the results are used for network design and investment planning, and also as the basis when expressing ISPs' opinions regarding Internet policy making.