[Mops] Fwd: [110attendees] Applied Networking Research Prize talks at IETF 110

Hi MOPS WG,

The talk by Francis Yan, in the IRTF Open Meeting on Monday at IETF 110, might be of interest – details below.

The recording of the talk is also at https://irtf.org/anrp/ <https://irtf.org/anrp/> if you can’t make the session. 

Colin

> Begin forwarded message:
> 
> From: Colin Perkins <csp@csperkins.org>
> Subject: [110attendees] Applied Networking Research Prize talks at IETF 110
> Date: 5 March 2021 at 19:19:45 GMT
> To: irtf-announce@irtf.org
> Cc: 110attendees@ietf.org
> 
> The Applied Networking Research Prize award talks for IETF 110 will be given by Audrey Randall <http://audrey-randall.github.io/> ("Trufflehunter: Cache Snooping Rare Domains at Large Public DNS Resolvers <https://doi.org/10.1145/3419394.3423640>") and Francis Y. Yan <https://francisyyan.org/> ("Learning in situ: a randomized experiment in video streaming <https://www.usenix.org/conference/nsdi20/presentation/yan>”).
> 
> Francis will talk about his work to apply machine learning to video bitrate adaptation. In streaming nearly four decades worth of video to more than 60,000 users across the global Internet, researchers found the statistical margins of error in quantifying streaming performance are considerable. These uncertainties affect the design space of machine-learning approaches that can practically be deployed. The team then developed an adaptive bitrate (ABR) algorithm that robustly outperformed other schemes by reinforcement learning in situ. To support further investigation, they are publishing data and results on the Puffer website and inviting the research community to train and test new streaming algorithms.
> 
> Audrey will discuss work on Trufflehunter, a DNS cache snooping tool for estimating the prevalence of rare and sensitive Internet applications. Unlike previous efforts that have focused on small, misconfigured open DNS resolvers, Trufflehunter models the complex behavior of large multi-layer distributed caching infrastructures. The large footprint of such resolvers presents an opportunity to observe rare domain usage, while preserving the privacy of the users accessing them. A controlled testbed allowed evaluation of how accurately Trufflehunter can estimate domain name usage across the U.S., providing a way to make a lower-bound estimate about the popularity of several rare and sensitive applications (most notably smartphone stalkerware) which are otherwise challenging to survey.
> 
> Look out for these talks in the IRTF Open Meeting on Monday – and find more information about the Applied Networking Research Prize at https://irtf.org/anrp/ <https://irtf.org/anrp/>
> 
> Colin Perkins
> IRTF Chair
> 
> -- 
> 110attendees mailing list
> 110attendees@ietf.org
> https://www.ietf.org/mailman/listinfo/110attendees