What Data Does Tor Collect on Pluggable Transports?

The Tor Project collects minimal, aggregated, and anonymized metrics regarding pluggable transport usage to monitor network health and track censorship events worldwide. To preserve user privacy, Tor bridges and relays process connection statistics locally, discarding identifiable details such as individual IP addresses before reporting aggregate figures to Tor Metrics. This article details the specific data points Tor collects, how user numbers are estimated, and the privacy safeguards applied to pluggable transport metrics.

Transport Protocol Types

Tor records which specific pluggable transport technology is being utilized. This includes protocols such as: * obfs4: Scrambles Tor traffic to look like random, unidentifiable data. * Snowflake: Routes traffic through temporary, volunteer-operated WebRTC proxies. * Meek: Uses domain fronting through major cloud CDNs. * WebTunnel: Mimics standard HTTPS encrypted web browsing.

Collecting this identifier helps developers understand which circumvention methods are most effective and which ones may be facing active blocking in specific regions.

Aggregated Country of Origin

When a user connects to a bridge using a pluggable transport, the bridge performs a local GeoIP lookup to identify the user’s country code. * The user’s individual IP address is discarded immediately after the country lookup. * Connections are grouped strictly by country code (e.g., ir, ru, cn, us). * To prevent individual tracking in regions with very low usage, connection counts below a specific threshold are combined or omitted to avoid de-anonymization.

Estimated User Numbers

Tor does not track distinct individuals or assign persistent user identifiers. Instead, user counts are mathematical estimates based on network activity: * Bridges measure the volume of bytes transferred or count the unique connection requests over a 24-hour window per transport and country. * The Tor Metrics system calculates the estimated daily active users by dividing total activity by the expected traffic volume of an average user session.

Network Protocol Distribution

Tor records the IP version used to connect to pluggable transports: * IPv4 vs. IPv6: Bridges report the proportion of incoming pluggable transport connections using IPv4 versus IPv6. This helps maintain compatibility and optimize network infrastructure across different internet architectures.

Data Tor Strictly Avoids Collecting

To maintain user anonymity and adhere to Tor’s core mission, the following data is never recorded, stored, or published: * Individual user IP addresses or MAC addresses. * Timestamps of specific user connections. * Destination websites, DNS queries, or browsing behavior. * Payload data transferred through the pluggable transport tunnels.

All collected data is published publicly on the Tor Metrics platform, providing researchers, human rights organizations, and network operators with transparency regarding global internet censorship trends without compromising individual security.