Papers, Presentations, and Regulator Filings
Papers, presentations, and other documents that describe the M-Lab platform, tests, and data, as well as research results obtained from the analysis of M-Lab data, are shown below.
Estimating Residential Broadband Capacity using Big Data from M-Lab
Knowing residential broadband capacity profiles across a population is of interest to both consumers and regulators who want to compare or audit performance of various broadband service offerings. Unfortunately, extracting broadband capacity from speed tests in public datasets like M-Lab is challenging because tests are indexed by client IP address which can be dynamic and/or obfuscated by NAT, and variable network conditions can affect measurements. This paper presents the first systematic effort to isolate households and extract their broadband capacity using 63 million speed test measurements recorded over a 12 month period in the M-Lab dataset. We first identify a key parameter, the correlation between measured speed and congestion count for a specific client IP address, as an indicator of whether the IP address represents a single house, or a plurality of houses that may be dynamically sharing addresses or be aggregated behind a NAT. We then validate our approach by comparing to ground truth taken from a few known houses, and at larger scale by checking internal consistency across ISPs and across months. Lastly, we present results that isolate households and estimate their broadband capacity based on measured data, and additionally reveal insights into the prevalence of NAT and variations in service capacity tiers across ISPs.
Access is more than cost: Measuring the quality of mobile broadband service
A4AI’s December 2018 report examines the policy dynamics that define the quality of mobile broadband service experienced by users across low- and middle-income countries (LMICs). Taking a step beyond our core research around affordable access and the cost of connectivity, this brief report informs ongoing efforts to translate Internet access into meaningful access for everyone.
Can Competition-Enhancing Regulation Bridge the Quality Divide in Internet Provision?
An analysis focusing on explaining differences in average connection speed per country per year using a pro-competition regulatory score constructed from country responses to an annual survey of national ICT regulators conducted by the International Telecommunication Union (ITU).
Reelección presidencial con los medios en crisis | Balance especial Elecciones 2018
An overview of the Internet conditions during the 2018 elections in Venezuela.
NETPerfTrace – Predicting Internet Path Dynamics and Performance with Machine Learning
We study the problem of predicting Internet path changes and path performance using traceroute measurements and machine learning models. Path changes are frequently linked to path inflation and performance degradation, therefore the relevance of the problem. We introduce NETPerfTrace, an Internet Path Tracking system to forecast path changes and path latency variations.
New Worldwide Broadband Speed League Unveiled
New data, extracted by Cable.co.uk from over 63m broadband speed tests worldwide reveals UK speeds are behind 30 other countries. At 16.51Mbps, the UK lags behind 19 European countries, 17 of them in the EU, but comes ahead of 158 other countries including Italy, France, Ireland and Monaco. Singapore ranks as the world’s fastest country with speeds of 55.13Mbps, with war-torn Yemen coming in last at an average speed of just 0.34Mbps
1,5 MB de lentitud y opacidad | Bitácora de las elecciones municipales del #10Dic
An overview of Internet speeds and access to information in Venezuela.
OECD Digital Economy Outlook 2017
The biennial OECD Digital Economy Outlook examines and documents evolutions and emerging opportunities and challenges in the digital economy. It highlights how OECD countries and partner economies are taking advantage of information and communication technologies (ICTs) and the Internet to meet their public policy objectives. Through comparative evidence, it informs policy makers of regulatory practices and policy options to help maximise the potential of the digital economy as a driver for innovation and inclusive growth.
The Record Route Option is an Option!
The IPv4 Record Route (RR) Option instructs routers to record their IP addresses in a packet. RR is subject to a nine hop limit and, traditionally, inconsistent support from routers. Recent changes in interdomain connectivity—the so-called “flattening Internet”—and new best practices for how routers should handle RR packets suggest that now is a good time to reassess the potential of the RR Option.
Strategic choice and broadband divergence in the transition to next generation networks: Evidence from Canada and the U.S.
This article investigates how infrastructure competition among broadband network infrastructure operators in Canada and the U.S. has influenced their incentives to increase fixed broadband connection speeds and invest in next generation fiber-to-the-premises (FTTP) technologies. The evolution of measured broadband speeds since the late 2000s documents growing differences in the incentives of dominant broadband operators to respond to demand for higher speed connectivity by increasing connectivity speeds they deliver to their customers. Dominant network operators in Canada have shown relatively stronger incentives than their counterparts in the U.S. to invest in and increase the capacity of legacy platforms. In the U.S. FTTP deployment incentives have been somewhat stronger, but network operators have been more reluctant to upgrade legacy technologies to deliver higher speeds. Diversity of strategic choices by large operators helps explain increasing regional and local broadband infrastructure gaps within the two countries. A high dividend payout financial strategy and increasing vertical integration appear to enhance the potential for overinvestment and inefficient duplication in legacy platforms by competing infrastructure providers.
Regulatory Federalism and Broadband Divergence: Implications of Invoking Europe in the Making of Canadian Telecom Policy
Invocations of Europe are increasingly evident in Canadian debates about wholesale broadband access, consumer protection, network neutrality and basic service policies. While some stakeholders in Canadian debates point to national and local approaches in Europe that appear to have been relatively successful in fostering broadband infrastructure development, others suggest that Europe is lagging behind and Canada should avoid Europe’s purported policy errors.
Towards estimating the untapped potential: a global malicious DDoS mean capacity estimate
What is the malicious reflected distributed denial of service (rDDoS) mean potential of the Internet? The authors have been using data from the openNTP project which measures the number of reflectors on the Internet since 2014 until now, and completed a graph that roughly estimates a lower boundary for global rDDoS mean potential across four Internet protocols (IPs); SSDP, NTP, SNMP and open recursive DNS.
Measuring Broadband in Schools
The challenges that school administrators face when budgeting for and deploying technology vary widely, as do their approaches to supporting its use within their schools. Measuring and assessing network health is a critical challenge facing public schools as they plan for both today’s and tomorrow’s broadband needs. New America’s Open Technology Institute (OTI) and Education Policy program partnered with the IT staff at Alexandria City Public Schools (ACPS) in Virginia to conduct a pilot study as a first step toward addressing this challenge. The study, Measuring Broadband in Schools, looked at the complexities of understanding network capacity in education institutions, and sought to better understand the challenges of measuring network capacity at the point of use in schools.
Measuring Latency Variation in the Internet
The authors analyse two complementary datasets to quantify the latency variation experienced by Internet end-users: (i) a large-scale active measurement dataset (from the Measurement Lab Network Diagnostic Tool) which shed light on long-term trends and regional differences; and (ii) passive measurement data from an access aggregation link which is used to analyse the edge links closest to the user. The analysis shows that variation in latency is both common and of significant magnitude, with two thirds of samples exceeding 100 ms of variation. The variation is seen within single connections as well as between connections to the same client. The distribution of experienced latency variation is heavy-tailed, with the most affected clients seeing an order of magnitude larger variation than the least affected. In addition, there are large differences between regions, both within and between continents. Despite consistent improvements in throughput, most regions show no reduction in latency variation over time, and in one region it even increases.
Analysis of Impartial Quality Measurements on Indian Broadband Connections
Broadband Internet user base is growing at an exponential pace in India. The market pressures are forcing the Internet service providers (ISPs) to sell their service offerings aggressively, resulting in a growing discrepancy between consumer expectations and service offerings from the ISPs. This discrepancy is often evident in experienced vs offered Internet connection service. This study examines the Quality of Service (QoS) variations for Internet connections of the Indian broadband users.
CUSUM Anomaly Detection
The CUSUM anomaly detection (CAD) method is based on CUSUM statistical process control charts. CAD is used to detect anomalous subsequences of a time series that show a subtle shift in the mean relative to the context of the sequence itself. CAD was applied in order to look for anomalies in MLab’s database of Network Diagnostic Tool (NDT) results.
Traffic Microbursts and their Effect on Internet Measurement
Report on M-Lab’s evaluation of reported site performance degradation in late 2015 due to switch discards caused by microburst traffic.
Measuring broadband performance using M-Lab: Why averages tell a poor tale
Broadband network performance is multi-faceted: it varies by ISP, by content source, by household connection, and by time-of-day. Daily or monthly averages, as published by content providers such as Netflix and Google, do not convey the full picture. In this paper we leverage M-Lab, the world’s largest open measurement platform, to characterize broadband performance across Australian households. Our study delves into millions of data samples collected from 96,882 households over four months, and looks beyond averages to make several interesting observations: 1) There is considerable variation amongst households, in terms of their broadband speeds and variability of network performance within a day and across days, and this information is lost when data is averaged across houses; 2) The fluctuations (even for a specific house) are significant, and can exhibit unexpected patterns, such as wide variations from one day to the next, and some clusters of outliers at certain times of the day. 3) By our experimental results, we conclude that neither aggregating by household nor aggregating by day or by hour is a sound measurement strategy. Moreover, our study sheds new perspectives on broadband evaluation by using M-Lab data, and can inspire future study into the underlying reasons of performance variation.
OECD Digital Economy Outlook 2015
This report assesses how countries can maximise the potential of the digital economy as a driver for innovation and inclusive growth, and discusses the evolutions in the digital economy that policy makers need to consider as well as the emerging challenges they need to address as a part of national digital strategies. Chapters include an overview of the current status and outlook of the digital economy; the main trends in the ICT sector, and developments in communication and regulation policy; and overviews of ICT demand and adoption, plus the effects of the digital economy on growth and development. This volume also includes a chapter on developments related to trust in the digital economy and on the emerging Internet of things.
Public QoS and Net Neutrality Measurements: Current Status and Challenges Toward Exploitable Results
The interest for Quality of Service (QoS) measurements in the Internet has exceeded the boundaries of research laboratories and passed into the public domain. The Internet is treated as a public utility, and users are given the right to access Internet content of their choice at an adequate quality. This coincides with the sustained interest for net neutrality (NN), basic safeguards for which are being taken in countries all over the world. Today, several tools have become available that empower end-users to measure the performance of their Internet connection and detect NN violations. However, the value that end-users obtain with these measurements is still small and the results are not being exploited satisfactorily by regulatory authorities, policy makers, and consumers. In this article, we perform a detailed review of the tools that are currently available for public QoS and NN measurements and explore three challenges that must be met in order to extract more value from the results: (a) Harmonization of measurement methodologies of basic performance parameters (throughput, delay, jitter, and packet loss); (b) the creation of a toolbox for detecting and monitoring NN violations; and (c) the use of a proper sampling plan for producing estimates over population groups.
Download link: http://www.jstor.org/stable/pdf/10.5325/jinfopoli.5.2015.0245.pdf
European capability for situational awareness
This study examines the field of Internet and Human Rights and more specifically, Internet Censorship measurement. The European Commission DG-Connect commissioned this study to assess the potential for a European Capability for Situational Awareness that continuously monitors and analyses the existence of Internet interference in combination with Human Rights Violations and recommend on potential implementation of such an ECSA. A proof-of-concept website was produced along with this report.
Analysis of End-User QoE in Community Networks
Community networks are a potential model for the Future Internet, where the users form and operate the network instead of a central, commercial entity. Socio-economic studies show that community networks are an excellent model to develop networking infrastructure commons (as common-pool resources or public goods) that promote sustainable development, with greater effects in less developed areas. The performance of parts of community networks has been studied extensively, often focusing on routing protocols or applications on top of community networks. This work focuses on the end-to-end quality of Internet access in community networks, as a validation of the technical applicability of this concept in under-served regions.
Measuring Broadband America Report - 2015
The FCC’s 2015 survey of Internet performance in the United States, with measurement performed against the M-Lab infrastructure
ISP Interconnection and Its Impact on Consumer Internet Performance—A Measurement Lab Consortium Technical Report
Research on the impact of ISP interconnections
- Study methodology referenced on page 35, formerly on code.google.com, may now be found on this website as a PDF.
- NDT source code and NDT Test methodology documents referenced on page 36, formerly on code.google.com, may now be found on GitHub: NDT Source Code - NDT Test Methodology
Challenges and Issues on Collecting and Analyzing Large Volumes of Network Data Measurements
New Trends in Databases and Information Systems 2014
Measuring Broadband America Report - 2014
The FCC’s 2014 survey of Internet performance in the United States, with measurement performed against the M-Lab infrastructure
Visualizing Internet-Measurements Data for Research Purposes: The NeuViz Data Visualization Tool
Overview of NeuViz, a network measurement visualization tool tailored to work with data from the NeuBot measurement test
Dimming the Internet: Detecting Throttling as a Mechanism of Censorship in Iran
Use of M-Lab’s NDT data to analyze Iranian Internet throttling and censorship
Strengthening Measurements from the Edges: Application-level Packet Loss Rate Estimation
ACM SIGCOMM Computer Communication Review 2013
Measuring Broadband America Report—February 2013
The FCC’s February 2013 survey of Internet performance in the United States, with measurement performed against the M-Lab infrastructure
Broadband Access to the Internet via Mobile Interfaces
Examination of mobile network performance indicators in the context of a European regulatory framework—IEEE WMCNT 2012
Network Neutrality and Consumer Discrimination: Comparing ISP’s GTCs and DPI Application
Examination of data from M-Lab’s Glasnost tool and other sources to explore its role in the network neutrality debate—EconStor 2012
Measuring Internet Performance When Broadband Is the New PSTN
Review of M-Lab and other sources of performance measurement data in an era when “broadband is the new PSTN”—MIT technical report 2012
Estimating Packet Loss Rate in the Access through Application-level Measurements
A proposed model for estimating packet loss rate by a given connection, using data from the Neubot measurement test
LIFEGUARD: Practical Repair of Persistent Route Failures
Examination of long-term network outages using data from measurement tests, and proposal of a system for detection, localization, and remediation
Unraveling the Economic and Political Drivers of Deep Packet Inspection
Identification and examination of deep packet inspection (DPI) used to throttle and block BitTorrent traffic, using data from M-Lab’s Glasnost test.
Measuring Broadband America Report - 2012
The FCC’s 2012 survey of Internet performance in the United States, with measurement performed against the M-Lab infrastructure
Broadband Internet Performance: A View from the Gateway
Presentation of the first study of network access link performance measured by BISmark by the architects of the BISmark router-based measurement test
Deep Packet Inspection and Bandwidth Management: Battles over BitTorrent in Canada and the United States
Examination of DPI use by Internet service providers to enable throttling or blocking of BitTorrent uploads or downloads—TPRC 2011
Measuring Broadband America Report - 2011
The FCC’s 2011 survey of Internet performance in the United States, with measurement performed against the M-Lab infrastructure
ShaperProbe: End-to-end Detection of ISP Traffic Shaping Using Active Methods
Methodology for detecting various types of traffic shaping—IMC 2011
Measurement Lab: Overview and an Invitation to the Research Community
Discussion of the scope of the effort by M-Lab founders, and an invitation to the research community to join them in expanding global measurement
Understanding Broadband Speed Measurements
Review of different speed measurement tests, including M-Lab’s NDT test, and examination of the benefits of different approaches
Discussion of the model for providing reverse trace information by the architects of the Reverse Traceroute test—NSDI 2010
Glasnost: Enabling End Users to Detect Trafﬁc Differentiation
Description of the rationale behind the Glasnost test and the way it can help expose opaque ISP network management by its architects—NSDI 2010
Pathdiag: Automated TCP Diagnosis
Review of the pathdiag test, used to diagnose network performance problems commonly aﬀecting TCP-based applications
Detecting BitTorrent Blocking
Review of the model used by architects of the Glasnost test to design a test to detect ISP throttling and blocking of the BitTorrent protocol
Additional Relevant Bibliographic References
The macroscopic behavior of the TCP congestion avoidance algorithm
ACM SIGCOMM Computer Communication Review; Volume 27 Issue 3, July 1997; Pages 67-82
The authors analyze a performance model for the TCP Congestion Avoidance algorithm. The model predicts the bandwidth of a sustained TCP connection subjected to light to moderate packet losses, such as loss caused by network congestion. It assumes that TCP avoids retransmission timeouts and always has sufficient receiver window and sender data. The model predicts the Congestion Avoidance performance of nearly all TCP implementations under restricted conditions and of TCP with Selective Acknowledgements over a much wider range of Internet conditions.We verify the model through both simulation and live Internet measurements. The simulations test several TCP implementations under a range of loss conditions and in environments with both drop-tail and RED queuing. The model is also compared to live Internet measurements using the TReno diagnostic and real TCP implementations. The authors also present several applications of the model to problems of bandwidth allocation in the Internet. They use the model to analyze networks with multiple congested gateways; this analysis shows strong agreement with prior work in this area. Finally, the authors present several important implications about the behavior of the Internet in the presence of high load from diverse user communities.
Government / Regulatory Filings
7/14/2017 - FCC: Restoring Internet Freedom, GN Docket No. 17-108
4/25/2017 - TRAI (India): Counter Comments on TRAI Consultation Paper on Net Neutrality
4/12/2017 - TRAI (India): TRAI Consultation Paper on Net Neutrality
10/11/2016 - NTIA: Comments on the National Broadband Research Agenda, Docket Number: 160831803-6803-01
2/19/2015 - FCC: Protecting and Promoting the Open Internet, GN Docket No. 14-28; Framework for Broadband Internet Services, GN Docket No. 10-127
1/10/2013 - FCC: Notice of Oral Ex Parte Communications, GN Docket No. 12-264
10/22/2012 - FCC: Inquiry Concerning the Deployment of Advanced Telecommunications Capability to All Americans in a Reasonable and Timely Fashion, and Possible Steps To Accelerate Such Deployment Pursuant to Section 706 of the Telecommunications Act of 1996, as Amended by the Broadband Data Improvement Act
9/21/2012 - FCC: Notice of Oral Ex Parte Communications, CG Docket No. 09-158, CC Docket No. 98-170, WC Docket No. 04-36
7/30/2012: Notice of Oral Ex Parte Communications, CG Docket No. 09-158, CC Docket No. 98-170, WC Docket No. 04-36
7/18/2012: Letter, CG Docket No. 09-158, CC Docket No. 98-170, WC Docket No. 04-36
6/29/2012: Notice of Ex Parte, CG Docket No. 09-158, CC Docket No. 98-170, WC Docket No. 04-36
6/29/2012: Amendment to Notice of Ex Parte, CG Docket No. 09-158, CC Docket No. 98-170, WC Docket No. 04-36
Experiences Monitoring Measurement Lab
PromCon 2017 Lightning Talk outlining M-lab’s transition to Prometheus monitoring.
Visualizing the Health of the Internet - Measurement Lab + Bocoup
Presented at the Open Data Science Conference in Boston, May 2017 Also presented at the Strata Data Conference in London, May 2017
M-Lab: What it is, why you should use it, and what’s happening in 2017
Presented at the Google Networking Research Summit, February 2017
M-Lab at IMC
Presented at IMC, November 2012
M-Lab—Network Science at Scale
Presented at MIT, November 2012
M-Lab—Open Internet Performance Data and How to Get it
Presented at the OECD—WPIIS, December 14, 2012
Other M-Lab Documentation
A brief overview of M-Lab
M-Lab Server Specifications for Contributors
A detailed technical review of the specifications and requirements for M-Lab server hosts
How to Get Involved with M-Lab
An overview of ways to join and support the M-Lab consortium
Requirements and Procedure for Approval of New M-Lab tests
A description of the requirements and process for approval of a new measurement test on the M-Lab platform
M-Lab Roles and Responsibilities
An outline of the roles and responsibilities for researchers using the M-Lab platform
M-Lab’s Founding Vision
The founding vision and operational procedures of the M-Lab measurement consortium, as laid out in 2009