Sponsor 2

Project Title: Defending Against Website Fingerprinting in Tor

Project Period: September 1, 2016 - August 31, 2019

Project Goals/Activities

The objective of this project is to engage in designing, extensive experimentation, and testing of recently proposed defenses against website fingerprinting attacks in Tor with the goal of near-term deployment. In particular, we are evaluating the WTF-PAD defense, which offers strong protection against the state-of-the-art attacks with no added latency costs and moderate bandwidth overheads that make it appropriate for use in Tor.

The WTF-PAD design also has potential defense applications against a wide variety of related traffic analysis attacks, which we are also investigating.

Project Tracking

Progress can be seen in the network team's kanban board filtering by sponsor 2.

Ticket Summary Status Owner Priority
#28634 Design a first useful padding machine (hiding client-side intro/rend circuits) merge_ready mikeperry Very High
#28638 Serialize state machines in the torrc new Low
#28694 When CircuitPadding is implemented in Tor, set it to 0 in sbws new Medium
#28777 Split circuitpadding.c code into multiple modules new Low
#28804 Add circuit padding to padding-spec.txt and write a doc for researchers new High
#28821 Introduce timer_is_scheduled() method and replace padding_scheduled_at_us new Low
#29084 Ensure circuit padding RTT estimate handes var cells/wide creates new Medium
#29102 Serialize padding state machine in consensus new Low
#29494 Optimize interaction between circuitmux and circuitpadding needs_revision mikeperry High
#29840 WTF-PAD: Log padding counters into the heartbeat new Low

Tickets already closed

Ticket Summary Status Owner Priority
#7028 Implement Adaptive Padding or some variant and measure overhead vs accuracy closed Medium
#22212 [warn] channelpadding_compute_time_until_pad_for_netflow(): Bug: Channel padding timeout scheduled 164729ms in the past. Did the monotonic clock just jump? closed arma Medium
#25501 Ensure WTF-Pad padding comes from the expected hop closed mikeperry Medium
#26632 Updated version of the WTF-pad proposal closed Medium
#26633 Solve histogram issues for WTF-Pad closed Medium
#26634 Enumerate unresolved issues that will need to be solved for a WTF-pad deployment closed Medium
#28142 Merge original WTF-PAD branch closed mikeperry Medium
#28610 will WTF-PAD impair bandwidth scanning? closed Medium
#28631 Release a useful WTF-PAD to the world (master ticket) closed Medium
#28632 Make the original WTF-PAD branch actually useful for us (submaster ticket) closed Medium
#28633 Pad on specific circuit purposes + ltups closed Medium
#28635 Clean up PADDING_TODO.txt closed Medium
#28636 Address WTF-PAD comments by Nick (2018-11-27 over IRC) closed High
#28637 Improve general WTF-PAD code quality and maintainability closed Medium
#28693 Add an option to disable circuit padding closed High
#28778 Rename some circuitpadding structs closed Medium
#28779 Remove double function declarations in circuitpadding.c/test_circuitpadding.c closed Medium
#28780 circpadding: Add machine flag for not closing circuit if machine is active closed Very High
#28878 WTF-PAD: Improve deterministic randomness in tests closed nickm Low
#29085 WTF-PAD: Reduce monotime usage because of performance issues closed Very High
#29169 32-bit warnings in test_circuitpadding.c closed nickm Medium
#29203 Add a way to specify machines as reduced circuit padding closed Medium
#29204 Inspect circuit queue during padding decisions closed Very High
#29298 Explicitly specify histogram bin endpoints/widths closed Very High
#29500 Broken circuitpadding unittests on appveyor closed mikeperry High
#29990 Handle zero monotonic time differences in the circuit padding code closed mikeperry Medium

Last modified 11 days ago Last modified on May 13, 2019, 7:06:47 PM