Markdown Version | Transcript | Session Recording | Session Materials

GAIA

Summary

The GAIA working group met at IETF 126 to discuss rural backhaul design, community network lifecycle modeling, human-centric methodology in networking research, open fiber data mapping tools, and operational practices for circular device management. The session featured five presentations, detailed architectural discussions on network sustainability, and updates on the group's adopted draft.

Key Discussion Points

1. Introduction and Agenda Review

• Slides: Agenda Slides
• Presenter: Kurtis Heimerl
• Discussion: Kurtis welcomed the participants to IETF 126, presented the standard Note Well guidelines, and adjusted the agenda to accommodate five speakers.

2. Next-Generation Wireless Backhaul Design for Rural Areas

• Slides: Next-Generation Wireless Backhaul Design for Rural Areas
• Presenter: Leonardo Maccari
• Technical Details:
  • Leonardo presented research on modeling the physical and economic feasibility of mesh-based wireless backhaul in underserved regions.
  • The study utilized 3D line-of-sight topographical models computed via GPUs across 10 rural Italian municipalities.
  • Evaluated equipment profiles included 5 GHz 802.11ac and 60 GHz 802.11ad devices, factoring in path loss models to calculate link capacity based on distance.
  • The economic model calculated CapEx (amortized over 5 years for wireless, 15 years for fiber) and OpEx (wholesale capacity, transport, maintenance, power) across various user penetration rates (25%, 50%, 100%). Results showed a monthly cost per subscriber ranging from €30 to €60, which was highly comparable to local DSL and satellite rates.
• Q&A:
  • Kurtis Heimerl questioned whether selecting the gateway node based on minimizing fiber-roll distance (the most expensive step) instead of topological centrality would yield better economic optimization. Leonardo noted that topological centrality reduces hop-count and link-capacity constraints, but optimizing purely for fiber length remains a valuable alternative.
  • Alistair Woodman observed that the estimated costs show why commercial telcos rationally ignore rural areas without receiving public subsidies, as density is too low to prove cost-effective.
  • Leandro Navarro asked how the market pricing has changed since the paper's initial data collection. Leonardo noted that hardware costs have decreased, but public infrastructure investments in these areas remain stagnant.
  • Kurtis Heimerl commented on the impact of Starlink on Wireless ISPs (WISPs), noting that many rural users choose Starlink over local WISPs due to perceived corporate reliability.
  • Bill Baraza asked about the specific hardware vendors simulated. Leonardo confirmed they used specifications modeled on Ubiquiti and MikroTik hardware planning platforms.

3. Community Networks Readiness and Maturity Model (CN-RMM)

• Slides:
  • Community Networks Readiness and Maturity Model
  • Marcelo:CN-RMM Community Network Readiness and Maturity Model
• Presenter: Marcelo Santos
• Technical Details:
  • Marcelo outlined the Community Network Readiness and Maturity Model (CN-RMM). The model addresses the operational "gap" where community networks fail after volunteers leave or hardware breaks due to lack of local governance.
  • The model maps community networks across five stages: community readiness (Stage 0), planning (Stage 1), deployment (Stage 2), operations (Stage 3), and evolution (Stage 4).
  • Rather than prescribing how things must be done, the model outlines what must be established (e.g., financial plans, diagnostic checklists, training records) to transition networks toward self-sustainability and transfer local technical control.
• Q&A:
  • Leandro Navarro emphasized that the model should remain non-prescriptive, observing that many rural communities already possess strong governance models (e.g., for managing water or forests) which they can apply to network infrastructure.
  • Leonardo Maccari highlighted overlapping work from the European netcommons.eu project, which created manuals and card games to introduce these exact organizational concepts to communities without bureaucratic friction.
  • Eduardo asked if the framework could offer a formal "certification" or stamp of maturity to help community networks access funding. Marcelo agreed that a certification model, perhaps trained or authorized by global entities like the Internet Society (ISOC) or regional NGOs, would be highly beneficial.

4. Human Methods in Networking Research

• Presenter: Kurtis Heimerl
• Technical Details:
  • Kurtis presented a paper published at HotNets 2024 advocating for the inclusion of qualitative, human-centered methodologies in systems and networking research.
  • He detailed three key methods: Participatory Action Research (PAR) (co-designing networks directly with stakeholders, such as opting for a distributed cell core over a cloud-hosted one for the Seattle Community Network because the community distrusted cloud providers); Ethnographic Methods (embedding in communities over long periods to understand informal dynamics, such as how BGP peering actually occurs over drinks at exchange points); and Positionality (researchers disclosing their backgrounds, financial backing, or activist leanings to contextualize their design decisions).
• Q&A and Chat Discussion:
  • Christopher Locke (ISOC Foundation) noted in the chat and on the mic that the Internet Society Foundation offers annual research grants for this type of human-centric research and emphasized that understanding usage patterns is fundamental to network sustainability.
  • Mohit Tahiliani shared in the chat that APNIC provides similar funding in the Asia Pacific region via the ISIF Asia grants.
  • Sophie supported the call to "reverse the gaze" and apply these ethnographic and positionality frameworks to corporate, for-profit networking structures and large Internet Exchange Points (IXPs).
  • Renzo Navas asked how researchers can find local partners and financing. Kurtis suggested incorporating these methodologies directly into standard systems research grants.
  • Rumesha noted the geographic bias in the systems community, where researchers working on networks in developing regions (such as Pakistan) are often forced by reviewers to prove how their work generalizes to the global North, whereas North-centric papers face no such standard.

5. Open Fiber Data Standard (OFDS)

• Presenter: Steve Song
• Technical Details:
  • Steve presented on the Open Fiber Data Standard (OFDS), an open-source structured data language designed to describe terrestrial fiber optic network layouts.
  • Because terrestrial operators often treat network maps as proprietary secrets, the OFDS project has built tools to crowdsource and standardize this geographic information.
  • He demonstrated a tool pipeline that:
    1. Imports static PNG/JPEG maps as image overlays into Google Earth to trace routes into KML files.
    2. Converts KML files into structured OFDS-compliant JSON and GeoJSON files while embedding metadata (owner, lease status, fiber strands, aerial/underground status).
    3. Uses "OFDS Studio", a custom plugin for the open-source QGIS application, to manage and visualizes these datasets.
    4. Automatically updates a global, collaborative, crowdsourced fiber map hosted on GitHub.
• Q&A and Chat Discussion:
  • Dan York shared a link to the OFDS Hackathon GitHub repository and highlighted the strength of the open-source QGIS ecosystem.
  • Kurtis Heimerl asked if QGIS acts as an open-source ArcGIS (which Dan York confirmed) and why the standard focuses exclusively on fiber. Steve replied that fiber is the foundation of middle-mile transit, but the standard will expand to cover towers, aerial passive infrastructure, and FTTx.
  • Leandro Navarro shared Spain's Guifi.net experience, noting that transparent fiber mapping was legally vital for protecting community investments from predatory claims by commercial operators.
  • John Levine raised the security concern of making physical fiber routes discoverable to bad actors. Steve replied that the risk is minimal compared to the benefit; of the 10,000 annual fiber cuts in countries like Nigeria and Ghana, almost all are accidental civil works mistakes that could be prevented by having publicly accessible maps.

6. Operational Practices for Digital Autonomy and Meaningful Connectivity through Circular Management of User and Network Devices

• Slides: Operational Practices for Digital Autonomy and Meaningful Connectivity through Circular Management of User and Network Devices
• Presenter: Leandro Navarro
• Technical Details:
  • Leandro discussed the recently adopted working group draft focused on circular device management.
  • The draft details the operations needed to sustain physical hardware in community networks, advocating for a "collective access model" where hardware (routers, switches, laptops) is managed as a community library commons.
  • Key requirements outlined in the draft include local repair training, secure data sanitization, components inventory management, and systematic e-waste repatriation to prevent toxic dumping in the global South. Leandro illustrated these concepts with case studies from Rosario and La Plata, Argentina.
• Q&A:
  • Kurtis Heimerl asked about the overlap with the newly formed Sustain group. Leandro acknowledged the overlap but argued that GAIA is the appropriate home because hardware circularity directly affects local community network deployment and operational autonomy.
  • Ben Roberts highlighted a major technical hurdle for recycled devices in Kenya and the wider global South: the upcoming end-of-life of Windows 10. He pointed out that most refurbished hardware cannot run Windows 11, and there is a critical shortage of local expertise regarding lightweight, long-term-supported Linux distributions. He requested that the draft or future research address lightweight OS selection and capacity building for flashing recycled machines.

Decisions and Action Items

• Circular Management Draft: Leandro Navarro will prepare an updated version of the adopted working group draft (draft-ietf-gaia-operational-practices) to resolve minor issues raised during the review, including alphabetizing definitions, fixing typos, and incorporating security/operational feedback.

Next Steps

• The working group will continue reviewing the circular device management draft on the mailing list.
• The chairs will coordinate with the Sustain group leadership during future IETF scheduling to prevent concurrent meeting times and minimize scheduling conflicts.

Related Documents

draft-ietf-gaia-operational-practices