Home
Glasnost Tests
Create your own test
Results
Contact
Glasnost has been designed to enable individual users to detect if their traffic is being shaped by ISPs. Using data gathered from individual users to infer throttling policies of ISPs is not straight-forward. First, we need to map users to their access ISPs. Second, we need a large sample of test users from each ISP. Third, the sample of test users should be representative (i.e., ideally a random sample) of the ISP's customer population.
Identifying the access ISP of a Glasnost user: First, we map the IP address of the user to the autonomous system (AS) she belongs to using data from the iPlane project. We then map the AS name to an ISP using a publicly available table from IANA.
Obtaining a large sample of test users for an ISP: Because tests by individual users have a small margin for error, we would need a large number of user tests from an ISP to accurately infer its policies. So we only included ISPs for which we have at least 100 user tests in the results presented below.
Obtaining test data from a representative (random) sample of users within an ISP: This is a much harder challenge because we do not actively recruit users to take Glasnost tests. Individual users run Glasnost tests on their own volition. It is possible that many of our testers are heavy users of BitTorrent that ISPs specifically target for shaping. So, for example, if 50% of Glasnost tests from an ISP showed throttling, it does not imply that 50% of users within the ISP are being throttled. However, measurement noise alone cannot explain why throttling was observed in such a large fraction (50%) of Glasnost tests, which suggests that the ISP must have deployed throttling for at least some of its users.
Thus, Glasnost data can be used to infer if an ISP is deploying traffic shaping — by checking if the percentage of Glasnost tests for an ISP that indicate traffic shaping far exceeds the measurement noise. However, Glasnost data cannot be used to accurately infer how widely traffic shaping is deployed within an ISP, i.e., the percentage of the ISP's users who are affected by traffic shaping.
If you have questions or feedback, you can contact us via e-mail:
