| RFC 9707 | BIAS Workshop Report | December 2024 |
| Kühlewind, et al. | Informational | [Page] |
The "Barriers to Internet Access of Services (BIAS)" workshop was convened by the Internet Architecture Board (IAB) from January 15-17, 2024 as a three-day online meeting. Based on the submitted position papers, the workshop covered three areas of interest: the role of Community Networks in Internet access of services, reports and comments on the observed digital divide, and measurements of censorship and censorship circumvention. This report summarizes the workshop's discussions and serves as a reference for reports on the current barriers to Internet access.¶
Note that this document is a report on the proceedings of the workshop. The views and positions documented in this report were expressed during the workshop by participants and do not necessarily reflect the IAB's views and positions.¶
This document is not an Internet Standards Track specification; it is published for informational purposes.¶
This document is a product of the Internet Architecture Board (IAB) and represents information that the IAB has deemed valuable to provide for permanent record. It represents the consensus of the Internet Architecture Board (IAB). Documents approved for publication by the IAB are not candidates for any level of Internet Standard; see Section 2 of RFC 7841.¶
Information about the current status of this document, any errata, and how to provide feedback on it may be obtained at https://www.rfc-editor.org/info/rfc9707.¶
Copyright (c) 2024 IETF Trust and the persons identified as the document authors. All rights reserved.¶
This document is subject to BCP 78 and the IETF Trust's Legal Provisions Relating to IETF Documents (https://trustee.ietf.org/license-info) in effect on the date of publication of this document. Please review these documents carefully, as they describe your rights and restrictions with respect to this document.¶
The Internet Architecture Board (IAB) holds occasional workshops designed to consider long-term issues and strategies for the Internet, and to suggest future directions for the Internet architecture. This long-term planning function of the IAB is complementary to the ongoing engineering efforts performed by working groups of the Internet Engineering Task Force (IETF).¶
The Internet as part of the critical infrastructure affects many aspects of our society significantly, although it impacts different parts of society differently. The Internet is an important tool for reaching the Sustainable Development Goals (SDGs) listed in [SDGs] and for globally supporting human rights. Consequently, the lack of meaningful access to digital infrastructure and services is also a form of disenfranchisement.¶
Solely having Internet access is not enough. At the same time as we work to connect the next billion people and reduce the digital divide, it is also important to understand persistent and novel inequalities in the digital age when accessing content and services. There are more and more barriers to meaningful access to the services and applications that run on the Internet. Even if Internet connectivity is available, information and service access may remain challenged and unequal.¶
This IAB workshop aimed to¶
collect reports about barriers to accessing content and services on the Internet. For example:¶
help the Internet community get a better understanding of how the Internet functions in different parts of the world and which technology or techniques need to be used to gain access to content.¶
build an understanding of what "being connected" to the Internet means: What is the Internet to users? What is needed to be meaningfully connected? What are the minimum requirements to be able to access certain parts of the content and services provided over the Internet?¶
This document is a report on the proceedings of the workshop. The views and positions documented in this report were expressed during the workshop by participants and do not necessarily reflect the IAB's views and positions.¶
Furthermore, the content of the report comes from presentations given by workshop participants and notes taken during the discussions, without interpretation or validation. Thus, the content of this report follows the flow and dialogue of the workshop but does not attempt to capture a consensus.¶
The workshop was organized across three days with all-group discussion slots, one per day. The following topic areas were identified, and the program committee organized paper submissions into three main themes for each of the three discussion slots. During each discussion, those papers were presented sequentially with open discussion held at the end of each day.¶
The first day of the workshop focused on the role of Community Networks [RFC7962] as a way to overcome the barriers to Internet access. Community Networks are self-organized networks wholly owned by the community and thus provide an alternative mechanism to bring connectivity and Internet services to those places that lack commercial interest.¶
Presentations ranged from highlighting the need for measuring Quality of Experience (QoE) for Community Networks, to the potential role the Content Delivery Network (CDN) can play in Community Networks, to the role of Satellite Networks, and finally, to the vital role of the spectrum in this space.¶
[MARTINEZ] highlighted the need to address QoE in discussions around Community Networks. As a community-driven deployment, the knowledge and involvement of individuals can vary; therefore, there are no guarantees of connectivity or quality of service. There is a need to focus on user expectations and how they translate to measurable performance indicators. Further, it asks for better documentation of best practices in deploying Community Networks as well as careful thought regarding manageability considerations for Community Networks in protocol development. [GUIFI] as an example Community Network was discussed, and some existing resources for Community Networks [APC] [ISOC] [TBB] were shared by the participants.¶
The inconsistent quality and performance of Satellite Internet result in a connectivity gap for Community Networks that rely on non-terrestrial networks (NTNs) for Internet access [HU].¶
[BENSON] focused on the prohibitive cost of transit and Internet services for Community Networks and argued for CDNs to provide transit-like and Internet services, at no more than at-cost, in a mutually beneficial way. Community Networks still need backhaul to and from the CDN's point of presence, and models for community-backhaul and open-source CDNs were highlighted. Discussion included the status of Project PANGEA [PANGEA] as well as legal and commercial considerations related to such use of CDNs.¶
[HU] highlighted that Satellite Internet provided by advanced low-Earth orbit (LEO) Satellite constellations can play a pivotal role in closing the connectivity gap in the urban-rural digital divide via Satellite-dependent Community Networks. These existing known performance and management gaps need to be focused on, to enable Satellite Internet to resolve the divide. Further, research directions such as multi-layer Satellite networking, autonomous maintenance, and integration between terrestrial networks and NTNs were suggested.¶
[RENNO] called attention to the coveted 6 GHz (part of the C-band with a desirable mix of coverage and capacity) as a prime choice for International Mobile Telecommunication (IMT) for 5G technology while it is in common unlicensed use in Community Networks (and small ISPs). Spectrum allocations directly impact industries and market access with ramifications for Community Networks. Further, there was a discussion on geopolitical tensions because of these allocations.¶
How can the technical community address the management gaps and improve best practices for Community Networks? Is the increasing complexity of the Internet making it more challenging to establish secure connections, and should this be taken into account in the design of the Internet? What steps need to be taken to make sure Community Networks are secure? Should manageability considerations be expanded to explicitly consider Community Networks? The Global Access to the Internet for All (GAIA) Research Group [GAIA] could be a venue for further discussion and research. Further discussion highlighted the need for readily available knowledge and tools for Community Networks as well as the tussle with market forces when commercial networks compete with Community Networks. Also, there is a lack of operational inputs from Community Network operators in the IETF/IRTF.¶
Critical Internet infrastructure affects many aspects of our society significantly, although it impacts different parts of society differently. The inequitable aspects are typically referred to as "digital inclusion"; these aspects signify that in efforts to digitalize society, there are those left out due to what is typically called the "digital divide", a related term specific to access to the Internet. These concepts together demonstrate that even if Internet connectivity is available, for some there will remain challenges towards achieving equality. This becomes especially significant as governments view the Internet as an important tool for helping them reach the SDGs listed in [SDGs] and for globally supporting human rights.¶
The second day of workshops was essential to understanding the nature of the digital divide. Presentations of reports interrogated at least three key aspects of the digital divide, though it is recognized that there may be more technical aspects of the digital divide that were not addressed. Three of those aspects were:¶
Ralph Holz presented research that exposes the more limited DNS-mediated access to government websites by Indigenous communities in Australia as compared to less disadvantaged users in the same population in "Evidence for a digital divide? Measuring DNS dependencies in the context of the indigenous population of Australia" [HOLZ]. DNS dependency trends were analyzed between two lists of domains serving Australian government sites for Indigenous users and the general population. Researchers found "evidence that dependencies for the indigenous population are indeed differently configured," indicative of a difference in service provisioning. However, qualitative follow-up research is needed to interrogate both the potential reasons for these differences and whether the differences contribute to a digital divide that is tangible for Indigenous users.¶
On the topic of availability of Internet services and content in multiple languages, "Universal Acceptance of Domain Names and Email Addresses: A Key to Digital Inclusion" was presented by Sarmad Hussain of ICANN [HUSSAIN]. The ICANN community has increased the options for multilingual identifiers through the expansion of the Internet's DNS for use in domains and email addresses. However, while the work of technical specifications and policy recommendations is complete, much work remains to deploy a multilingualized Internet. Today, there are around 150 internationalized domain names (IDNs), but equal rollout of these scripts at the domain level is hindered primarily by software and applications that do not yet recognize these new scripts. "Universal Acceptance" is a program of action for the Internet community at large that can ensure that IDNs are accepted and treated consistently.¶
In "A Framework for Improving Web Affordability and Inclusiveness", Rumaisa Habib presented research on the connection between website size and cost to end users [HABIB]. This critical inquiry presents access in terms of affordability and through measurement demonstrates that the material costs to end users who pay for their connection based on the volume of data they download and upload have risen as the complexity of the Web grows. Their research provides a framework for optimization based on end-user affordability. This framework is anchored in reality: [HABIB] proposes a fairness metric and suggests systematic adaptations to complexity of the Web based on "geographic variations in mobile broadband prices and income levels."¶
These three reports discuss very different aspects of current inequalities in Internet access in various parts of the world: service provisioning, availability, and economic costs. Notably, the reports discuss trends that exacerbate the digital divide beyond the question of connectivity or whether users have access to the Internet, potentially yielding concrete ways that the IETF community can address digital inclusion within its remit.¶
Discussants noted that while there are some interesting aspects to the problem of the digital divide, such as measurements and frameworks, most of the work involves getting this work to the appropriate people at the policy level; therefore, it is important to communicate this technical evidence to the appropriate people. The IETF's role could be to build consensus on the proper solutions presented to decision-makers that put research and measurement not only in context but also in a consensus-driven solution space. Another method to better communicate this research is by telling stories of end users in more relatable and relevant terms; this is often a challenge at the technical level and a role for more diverse stakeholders at the more local level.¶
This session focused on reports of censorship as observed in recent years in different parts of the world; it also focused on the use of, and expectations for, censorship circumvention tools, mainly the use of secure VPN services.¶
The censorship reports highlighted legal frameworks and court actions that put legal obligations on regional network providers to block traffic. The discussion focused on Asia, specifically India, and included Russia as an example where censorship practices have recently undergone significant changes. Further, measurements to validate the blocking as well as analyses of how blocking is implemented were discussed, i.e., which protocols are used but also which kinds of devices are used to configure the blocking rules and where they are deployed.¶
[SAMSUDIN] reported on confirmed blocking from 10 countries (Cambodia, Hong Kong (China), India, Indonesia, Malaysia, Myanmar, Philippines, Thailand, Timor-Leste, and Vietnam) in the period from 1 July 2022 to 30 June 2023. The blocking was confirmed by either (1) Open Observatory of Network Interference (OONI) measurements for existing blocking fingerprints or heuristics (i.e., for new blocking fingerprints as well as news reports of blocking orders) or (2) user experiences. Most of these countries block specific content such as porn, gambling, or certain news pages. Interestingly, the blocking in Hong Kong and Myanmar is focused on the military and governmental pages of foreign countries. Blocking is often realized by either DNS tampering or HTTP tampering. For DNS, either a chosen IP address, a bogon IP address (e.g., 127.0.0.1), or an empty domain (NXDOMAIN) is used. In the case of DNS tampering using a chosen IP address or HTTP tampering, some countries provide a block page that exposes the blocking; however, more transparency related to blocking is requested by civil society organizations and the Internet Monitoring Action Project (iMAP).¶
[GROVER] further focused the discussion on online censorship in India, Pakistan, and Indonesia.¶
As discussed in [Singh2020], in India, where providers are responsible for implementing the blocking but no method is mandated, the six major ISPs (covering 98.82% of all subscribers) were tested on a total of 4379 blocked websites (based on court orders, user reports, and publicly available or leaked government orders) by using DNS poisoning/injection or using censorship based on HTTP or the Server Name Indication (SNI). The censorship techniques used and websites blocked were different across ISPs. Multiple ISPs used two different techniques (depending on the website), and all but one provided censorship notices. A list of 4379 potentially blocked websites was tested; 4033 of those websites appeared in at least one ISP's blocklist. Providers blocked between 1892 and 3721 of the 4033 websites, with only 1115 websites (27.64%) blocked by all six ISPs.¶
In contrast, in Pakistan, the government can also order the ISPs to perform blocking, and blocking has even been observed in the past at the Internet Exchange Point (IXP) level. Since 2020, there has also been a central Web Monitoring System deployed at lines of international connectivity. In Indonesia, initially, the government guided ISPs in how to perform the blocking. The regulations were updated in 2020 to allow Indonesian ISPs to block websites at their discretion. In 2022, there was a proposal by ISPs to centralize DNS. In Indonesia, a partial blocklist is publicly available, but without any indication of why something is blocked [Grover2023].¶
[BASSO] reported that for Russia a large increase in additions to Roskomnadzor's blocklist was observed in March 2022 as well as in December 2022, foremost covering news pages but also covering human rights organizations and social media, where more than 3500 blocking orders were added to the list by an "Unknown body". Further, blocking of domains that are not in the official Roskomnadzor blocklist has been observed as well.¶
An invited talk included a presentation of the work in [WANG] on locating censorship devices by using HTTP and TLS traceroutes, identifying device vendors through fingerprinting, and reverse-engineering censorship triggers by the use of fuzzing. For example, in the case of Azerbaijan and Kazakhstan, they showed that a significant portion of measurements from remote countries is blocked at the endpoint, indicating local policies, but showed that connection resets are also happening in Belarus and Russia. Further, they could identify a set of commercial network devices (with filtering techniques such as firewalls) that are used in these countries for censorship and show how fuzzing can be used to fingerprint and cluster behaviors as well as potentially circumvent the deployed methods.¶
All speakers called for more transparency by requiring blocking messages as well as publication and auditing of blocklists. Potentially, even standardization could help.¶
Further on in the session, the possibility and prevalence of using VPNs for circumvention were discussed, including user expectations and an analysis of security shortcomings of commercial VPN services. The analysis presented in [RAMESH-1] has shown various problems that lead to data leaks, such as (1) leakage of IPv6 traffic, (2) non-browser traffic, or (3) tunnel failure, in addition to failing to uphold user expectations, especially when used in authoritarian regimes for censorship circumvention or private communication.¶
The question of how common the use of VPNs for circumvention is and its legal implications, as VPNs are illegal in a few countries, was discussed. For example, VPNs are not officially banned in India, but VPN providers need to store log data and those who haven't complied stopped serving India. However, more data on VPN use and blocking might be needed.¶
After all, there is a cat-and-mouse game between censorship and circumvention; however, continued work on protocol enhancements that protect user privacy is essential.¶
Some key takeaways from the workshop are as follows:¶
There is a need for the technical community to address the management gaps in operating Community Networks.¶
Work should be done to document best practices for operating Community Networks.¶
During the development of protocols, explicit manageability considerations related to Community Networks should be carefully thought out.¶
Build consensus on solutions that have the most significant impact in fostering digital inclusion. Further, promoting these solutions ensures that efforts to bridge the digital divide are effective and inclusive.¶
Further work should be done to enhance protocols, ensuring that user privacy is preserved.¶
Develop further protocols (or extensions to existing protocols) that enable more transparency on filtering, and promote their use and deployment.¶
Develop new VPN-like services and potentially support measurements to understand their deployment and use.¶
Further discussion of these topics could happen in the GAIA Research Group, the Human Rights Protocol Considerations (HRPC) Research Group, the Privacy Enhancements and Assessments Research Group (PEARG), and the Measurement and Analysis for Protocols Research Group (MAPRG), based on relevance to the research group. Management-related and operations-related discussions can be taken to the IETF Operations and Management Area Working Group (OPSAWG). The community could also explore whether a group focused on censorship (and its circumvention) could be created.¶
This document has no IANA actions.¶
This document has no security considerations.¶
Nineteen position papers were submitted to the workshop call for papers. Twelve were selected for publication. Papers that were not published either (1) only provided a very prelimited analysis of an idea that was felt to be incomprehensive for discussion at the workshop or (2) addressed problems that were considered beyond the scope of the workshop discussions, e.g., discussing cybersecurity threats as a barrier to participation or implication of technology in a regulation that imposes blocking. Both of these scenarios pose a potentially severe risk for the open Internet; however, they might pose a high risk for all Internet users but do not necessarily imply an unbalance.¶
All accepted papers are available at <https://datatracker.ietf.org/group/biasws/materials/>.¶
This is the list of all published papers:¶
Community Networks:¶
Martínez-Cervantes, L. M. and R. Guevara-Martínez: Community Networks and the Quest for Quality [MARTINEZ]¶
Benson, T. A. and M. Fayed: A 'C' in CDN: Access service to and from the Internet at cost for community networks [BENSON]¶
Hu, P.: Closing the Performance and Management Gaps with Satellite Internet: Challenges, Approaches, and Future Directions [HU]¶
Rennó, R.: Maximising Connectivity: The Spectrum's Vital Role in Technology Access [RENNO]¶
Digital Divide:¶
Holz, R., Nazemi, N., Tavallaie, O., and A. Y. Zomaya: Evidence for a digital divide? Measuring DNS dependencies in the context of the indigenous population of Australia [HOLZ]¶
Hussain, S.: Universal Acceptance of Domain Names and Email Addresses: A Key to Digital Inclusion [HUSSAIN]¶
Habib, R., Tanveer, S., Inam, A., Ahmed, H., Ali, A., Uzmi, Z. A., Qazi, Z. A., and I. A. Qazi: A Framework for Improving Web Affordability and Inclusiveness [HABIB]¶
Ott, J., Bartolomeo, G., Bese, M.M., Bose, R., Bosk, M., Guzman, D., Kärkkäinen, L., Kosek, M., Mohan, N., Trossen, D., Welzl, M., and L. Vogel: The Internet: Only for the Fast? [OTT]¶
Ohlsen, L.Y.: BIAS workshop - M-Lab Position Paper submission [OHLSEN]¶
Censorship:¶
Samsudin, S.: iMAP (Internet Monitoring Action Project) 2023 Internet Censorship Report [SAMSUDIN]¶
Grover, G.: The infrastructure of censorship in Asia [Grover2023]¶
Basso, S.: How Internet censorship changed in Russia during the 1st year of military conflict in Ukraine [BASSO]¶
In addition to the submitted papers, two invited talks were included, based on published papers:¶
The workshop participants were Arnaud Taddei, Carlos Pignataro, Carsten Bormann, Cindy Morgan, Colin Perkins, Cory Myers, Dan Sexton, David Guzman, David Millman, David Schinazi, Dhruv Dhody, Gurshabad Grover, Hanna Kreitem, Jane Coffin, Jiankang Yao, Jörg Ott, Juan Peirano, Lai Yi Ohlsen, Luis Martinez, Mallory Knodel, Marwan Fayed, Matthew Bocci, Michael Welzl, Michuki Mwangi, Mirja Kühlewind, Mona Wang, Peng Hu, Ralph Holz, Raquel Rennó, Reethika Ramesh, Rumaisa Habib, Sarmad Hussain, Simone Basso, Siti Nurliza Samsudin, Suresh Krishnan, Theophilus Benson, Tirumaleswar Reddy.K, Tommy Pauly, Vesna Manojlovic, and Wes Hardaker.¶
The workshop program committee members were Christopher Wood (IAB, Cloudflare), Dhruv Dhody (IAB, Huawei), Mallory Knodel (IAB, NYU), Mirja Kühlewind (IAB, Ericsson), and Tommy Pauly (IAB, Apple).¶
Internet Architecture Board members at the time this document was approved for publication were:¶
Thanks to Arnaud Taddei for helpful suggestions to improve this report.¶