NNTools 2015

Multidisciplinary Workshop on Network-Measurement Methodologies and Tools Applied to Net Neutrality

Abstract

NNTools2015 puts together people working on experiments and people that do policy in the network neutrality domain, so that they can learn from each other. People working on measurements should help others to better understand the Internet, what could be measured, and what could be interesting research questions given what could be measured. Conversely, social scientists and people working on policy should help people working on network measurements to understand their policy/research questions, and to understand which experiments could be helpful to answer such questions.

Info on the Workshop

The workshop, organized in the context of EINS, the Network of Excellence in InterNet Science, focuses on the intersection of Internet measurements and policy. It will take place in Turin, at the Nexa Center (Via Boggio 65/a, Torino). There will be an afternoon session on the 19th of January and a morning session on the 20th.

The aforementioned sessions are going to be a workshop in the strictest sense: it will include members of the EINS Network and external invited participants, but it is not going to be open to the public. All participants will be already familiar with the discussed issues and we plan to focus on the interaction between peoples doing tools/measures and people expert in the policy/policy support arena. The following are some basic ideas to frame the workshop:

Framework: We understand Network Neutrality in its broadest, abstract meaning, and we use such concept as a benchmark (without any normative implication): packets in transit are all treated alike, regardless of their source, protocol, content, or other characteristics. It is likely that no network will be perfectly neutral: our only assumption is that measuring deviations from this benchmark is frequently interesting. Various kinds of (passive and active) measures may be of interest, and both measures from "within" the network and from its "edges" are relevant.

Opportunity: There is room to improve the communication and interaction between those who develop tools and perform measurements (i.e., those who try to better understand/make more transparent the Net) and those who deal with policy (or business) (i.e., those who consider the Net at a higher level of abstraction).

Method: This workshop aims at establishing a direct dialogue between those who deal with network measurements/tools and those who those who deal with policy. We aim at providing reading material beforehand and having just short presentations, leaving as much room as possible for discussion.

Objective: Our goal is to understand which tests could better inform policy making, and which characterizations of the Network are already available, but possibly neglected in policy reasoning.

Topics: During the preparation of the workshop, we will welcome suggestions concerning relevant topics. To date, we consider that the following topics are surely relevant: issues concerning the interaction between providers (inter-domain congestion, de-peering); interferences with users’ traffic (protocol shaping understood in a broad sense and including any throttling, or content blocking).

Expected outcomes: Some expected outcomes (to be better defined in collaborations with the participants) include: mapping of existing empirical evidence (and how to find it); identifying TODOs related with making it easier to collect/process such evidence; mapping of missing empirical evidence (both in terms of useful and "easily" implementable tools, and in terms of hardly implementable ones, i.e., the challenge for hackers).

Hacking: The day before the workshop (Sunday 18 January) and the morning of the 19 January; the afternoon of the 20 January and the 21 January (TBC). Possible topics: Neubot, OONI, Libight, M-Lab and M-Lab data (other suggestions from interested developers are welcome).

Program

19 January 2015

Workshop Session #1 “Measurements -> Policy”: 14:30 - 18:30

The objective of this session is to discuss existing measurement tools to highlight their strengths and weaknesses. This is functional to discuss how to improve the state of the art among people working on tools and, more importantly, to explain people working on policy what could and what could not be measured easily.

14:30 - 14:45: Welcome and framework (Juan Carlos De Martin and Simone Basso)
14:45 - 15:15: Round table: speak 2’ to introduce yourself and why you’re here
15:15 - 16:15 Lightning talks

• Simone Basso, Nexa Center at Politecnico di Torino, Neubot project [Slides]
  I will describe Neubot, a network measurement tool for running network-performance measurements from the edges developed here at Nexa Center. I will discuss the results of an ongoing experiment which seeks to compare the results of network-performance measurements emulating HTTP and BitTorrent. I will explain what policy conclusions can be extracted from that data, and what is missing to make stronger policy conclusions.
• Collin Anderson, M-Lab [Slides]
  While interconnection is a core principle of the Internet, the role of interconnection health and traffic management practices in end consumer performance has remained opaque to users, policymakers, and researchers alike, based on the complexities of conducting measurement at scale. Measurement Lab’s research on interconnection performance has exposed episodes of persistent congestion across multiple Access ISPs and Transit ISPs over the past three years. These incidents lead to substantial degradation of end-user connectivity that have been quantified and detailed through the substantial performance data made available through M-Lab. Due to time limits of the lightning talk, we will focus on the technical findings of the study and discussing the toolkit released concurrently with the report in order to elaborate on the open datasets and tools that M-Lab has made available to interested researchers.
• Arturo Filastò, HERMES Center for Transparency and Digital Human Rights, GlobaLeaks, The Tor and OONI projects [Slides]
  I will give a short overview of the OONI project. I will illustrate what are our guiding principles and differences with different other existing project. Some examples of data that has been collected with OONI and what are the conclusions and social impact achieved thanks to them.
• Enrico Gregori, Istituto di Informatica e Telematica - Italian National Research Council and Valerio Luconi, Portolan project [Slides]
  The incompleteness of the Internet AS-level topology is the most serious pitfall of research studies concerning the Internet inter-domain ecosystem. In this talk we show two projects, Portolan and Isolario, aimed at filling this gap by offering services to different type of users. Portolan is a smartphone-based crowdsourcing system which provides an app for Android that volunteers can install to run traceroute measurements to discover the Internet topology. Isolario provides real-time services for network operators in exchange of their full routing table.
• Marco Mellia, Politecnico di Torino, mPlane and Tstat projects [Slides]
  Privacy tangle: HTTPS and third party trackers. Who is going to protect my daughter?
  Increased user concern over security and privacy on the Internet has led to widespread adoption of HTTPS, the secure version of HTTP. HTTPS authenticates the communicating end points and provides confidentiality for the ensuing com- munication. However, as with any security solution, it does not come for free. Given the opaqueness of the encrypted communication, any in-network value added services requiring visibility into application layer content, such as caches and virus scanners, become ineffective.
  At the same time, during the visit to any website, the average internaut may face so called online trackers, invisible third party services that collect information about users and profile them. We list more than 400 tracking services, of which the top 100 are regularly reached by more than 50% of Internauts, with top three that are practically impossible to escape. More than 80% of users gets in touch the first tracker within 1 second after starting navigating. And we see a lot of websites that hosts hundreds of tracking services.
  Unsurprisingly, tracking services are also embracing HTTPS, so that it becomes almost impossible to check which information they collect.
  In this tangle, who is going to protect us?
• Walter De Donato, Università di Napoli, HoBBIT project [Slides]
  The talk will give a brief overview of what we learned while designing and operating the HoBBIT platform, which enables fully configurable and extensible management and coordination of distributed network measurements from portable software agents running on user devices.

16:15 - 17:15 Discussion after the lightning talks
17:15 - 18:15 Break out session (ISP interconnection / shaping and throttling)

• Detection of blocking and shaping of protocols, and its impact on policy
• Detection of ISPs interconnection issues (e.g., Comcast / Level3), and its impact on policy
• Privacy and online tracking

18:15 - 18:30 Wrap up of break out session and conclusions

20:00 Social Dinner at "il Circolo dei Lettori", (Palazzo Graneri della Roccia) Via Giambattista Bogino, 9, Torino (map)

20 January 2015

Workshop Session #2 “Policy -> Measurements”: 09:30 - 13:30

The objective of this session is to discuss what evidence is needed to guide policy research on network neutrality, and to identify how existing tools can be used and/or improved to produce meaningful data useful to move policy research forward, and/or how existing datasets collected by existing tools could be combined to achieve the same goal.

09:30 - 10:00: Lightning talks

• Juan Carlos De Martin, Nexa Center, EINS project
• Chris Marsden, University of Sussex, EINS project
• Bendert Zevenbergen, Oxford Internet Institute
  Ben is developing a policy paper about the value of Internet measurements to provide an evidence base for Internet policy making. This paper will be presented to policy makers during a dedicated workshop at the European Parliament in May 2015. It aims to address the interdisciplinary gap between engineers and policy makers by explaining in layman’s terms what the benefits (and limitations) of Internet measurements can be for evidence-based policy making. The paper will explore (1) how the Internet measurements field can provide data and analysis for current policy debates, (2) how an interdisciplinary approach can uncover new research questions that could impact policy debates, and (3) how this interdisciplinary approach can also uncover currently unexplored situations where new policies may be needed. For a multi-disciplinary approach, Ben is seeking several co-authors from different fields.
• Ben Wagner, European University Viadrina [remotely from Berlin]
• Luca Cicchelli, TOP-IX Consortium [Slides]
  IXPs are key points of current Internet architecture. Even if ISPs have many systems to discriminate Internet traffic in their internal networks they often act at interconnection level, especially with content providers. TOP-IX experience on this framework will be showed in the talk.
• Kavé Salamatian, Université de Savoie [Slides]

10:00 - 11:00: Discussion after the lightning talks
11:00 - 11:15: Coffee break
11:15 - 12:45: Bringing all together and preparing a short paper for the Internet Policy Review open-access journal (lead: Federico Morando, Nexa Center at Politecnico di Torino, and Xin Wang, University of Southampton)
12:45 - 13:00: Goodbyes

Images from the workshop are available on Flickr.