[katzenpost] mixnet notes from visit to KU Leuven

[dawuud]

Hello mixnet people.

Here's some notes of the discussions we had about mixnet implementations
with Claudia and Tariq at KU Leuven on February 5th (yesterday):


* Mixnet researchers use simulations to calculate entropy, average mix
  anonymity set size and so on.  Instead of using a simulation which
  may well have different AQMs than the katzenpost implementation.
  We could instead perform emulation tests. Currently, we use a tool
  called kimchi which runs an entire katzenpost mix network in a single process.

  Tariq: questions for you:
  
  1. What format is your test packet log written in?
  2. What kind of statistical output should be produced by such a mixnet emulator?
  3. What other features should such an emulator have?
     Which Loopix parameters will you want to specify?
     Currently we have not implemented any decoy traffic (dummy messages).
     Do you require decoy traffic (Loopix design) in order to
     perform meaningful tests with our emulator?
     We can implement essentially any features you want so please elaborate.

  We would like to make a useful mixnet emulator that you can use based on
  https://github.com/katzenpost/demotools/tree/master/kimchi

* When should mixes start counting the poisson mix delay?

  There's two ways to do this:

  1. what we do now: start the poisson counter the moment the packet is received.
  
  2. an alternate implementation: start counting the poisson mix delay
     when the packet arrives in the mix queue scheduler. Our
     retransmission timer is set to:

    timer_duration = Poisson_RTT + send_slack

    where send_slack is some constant value to account for additional
    network and processing delays. We might increase send_slack to account
    for the increase in variance.

* future research: explore performance and information leakage tradeoffs for
  message fragmentation schemes involving two or more packet sizes.

* if a mix's egress queue is backlogged can this cause the mix strategy and
  ingress queues to also become backlogged?

     * "this will likely never happen... unless an upstream adversary
       is dropping the mix's egress packets causing TCP to retransmit
       and make slow forward progress.

* Yawning Angel has clarified some AQM implementation details for us in these two tickets:

  * https://github.com/katzenpost/server/issues/46

  * https://github.com/katzenpost/server/issues/47


Have I missed anything?
Please do reply if this e-mail beckons you to.


Sincerely,

David Stainton
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: not available
URL: <https://lists.mixnetworks.org/pipermail/katzenpost/attachments/20180206/3935f4f9/attachment.sig>