Minutes IETF114: gaia: Wed 10:00
|Meeting Minutes||Global Access to the Internet for All (gaia) RG|
|Title||Minutes IETF114: gaia: Wed 10:00|
Minutes GAIA group at IETF 114
- Date: 2022-07-27 (Wednesday)
- Time: 14:00 UTC (GMT/UTC) 120 mins
- Documents: https://datatracker.ietf.org/meeting/114/session/gaia
Agenda Bashing, Minutes taker, Blue Sheets, etc. (10 min)
Introduction of a new co-chair, Kurtis Heimerl, welcome! Leandro will
plan to step down after this meeting but still participate.
Satellite-integrated community networks
Peng Hu, University of Waterloo
Starlink and OneWeb constellations, current low-earth-orbit satellite
Internet providers. have global coverage.
Satellite platforms driven by decreased costs of satellite launches;
potentially very small/light satellites (shoebox-size)
Some new satellite platforms proposed, with high bandwidth.
Satellite a key option for community networks on global scale; has the
most coverage possible.
3gpp considering integration of leo satellites into 5G.
Notes that some organizations like ISOC have LEOs in their action plan
with respect to Community Networks (CNs).
Presenter notes that LEOs could be part of an integrated CN systems and
are called SICNs in this presentation.
Network management gap is key to address, and presentation posits that
autonomous maintenance is critical, a hierarchical approach to
self-maintenance should be part of SITNs.
Introduced a case study with Mininet VMs ith FRRouting stack for BGP and
Zebra dump parser. Results show that promising ML algorithms are
available. Also performance improvements for a hierarchical approach.
Conclusions SICNs provide possible future options for community
Q From Rajeev: Noting the "coverage map" on Slide 8, it seems overly
sparse for dense areas of the US. A: Graph is based on specific data and
Q from Kurtis: Failure does not required ML, usually it is the power. We
are going into the community to run courses about the technology. How
does this fit into your vision of creating resilient network? A: Lots of
different networks with lots of different failure modes. Some can be
helped by ML. No "one-size-fits-all" solution to resiliance.
Q from John: Are you concerned about "space junk" or is that out of
scope? A: Not out of scope, very concerning.
Q from Nabil: You used mininet for simulation. It has scaling issues.
Could use Linux namespace instead. A: Thanks for the info.
Resilient Education Information Infrastructure for the New Normal (REIINN) project
Philip A. Martinez, Philippines' Department of Science and Technology
Project REIINN. Lots of areas missing Internet connectivity which
doesn't allow for remote learning during COVID. Use UHF to provide LTE
service to remote areas and provide education content. LokaLTE
architecture is described. Broadcast educational content through
specialized UEs "set top box". Implementation challenges discussed
including permits, import restrictions, and redunancy.
Questions/comments include how to connect and share learnings.
Kurtis asked about how to provide information and education on how to
join the internet, how the technology works and issues like what is fake
N50 Digital Inclusion Pilot @Luumbo, Zambia
R. K. Rajeev, picoNETS N50 Partner & Kevin Schwartz,
GeeksWithoutFrontiers N50 Lead Partner
Kevin: n50 project is a consortium of organizations focused on the usage
gap -- there are many barriers, not just access, but lack of
applications or costs. Demonstrate novel, replicable solutions that
include services that support the adoption of broadband; models that
communities can replicate with other partners.
Curating projects, and matching partners as needed for each project and
Pilot project in Luumbo, Zambia: 4000 residents, mostly subsistence
Rajeev: factors including costs, digital literacy. pilot project in a
small-knit community so that we can clearly see the impact of
interventions. working with a telco in that area, which was currently 2g
or spotty 3g. observed how the community responded to the rollout of 4g
Seeding the community with 4g capable devices, in particular the school
or healthcare. Collecting usage data on trends.
Next, the content side: with access and devices, what will the people
actually consume? want to provide specific, relevant content that is
accessible at a reasonable cost.
AMN provided a 4g tower, with a satellite backhaul which is a
constrained network. Made a replicable stack for the local tower:
including radio, CDN and caching, content providers. Education is
important; but also a commercial FM station which has financial and
agricultural content; religious content; health. Keeping costs low and
technology simple so that adding a small site is feasible for a telco.
Reporting on how digital utilization grows for different interventions
over a 2+ year project. Goal is a sustained improvement in digital
literacy and engagement.
Highly replicable existing tech stack, a template that should be
deployable in any small community.
Applications are relevant to the community and to the schools, who have
access to lessons/materials they didn't before. Health clinic and remote
health services has increased timeliness of health services.
Sustainability plan in place from the start; needs a business model that
will be worthwhile for businesses to continue to invest. Committed to
exit strategies, n50 doesn't want to be a provider for the long term, or
create a dependency situation.
Whitepaper, and hearing interest in repeating this model in other
settings, with other content focus (e.g. agriculture):
Leandro: technology can bring benefits but also risks. how does it
change power structure in the community? new jobs in maintaining
infrastructure? have you tried independent analysis of the impact to
ensure that the overall impact is positive?
Kevin: k12 adoption of computers has demonstrated good and bad. but
communities that don't have access are increasingly marginalized because
other communities are adopting; disconnected communities are already
being negatively effected. try to be purposeful about applying it in the
way that the community wants. [...] currently focusing on the positive
outcomes that we can get.
Rajeev: early step was community liaison, prior to technical or
measurement work. unless the community feels involved in both planning
and execution, there won't be buy-in on the potential of the project.
avoiding "helicopter" approach that drops in devices. multiple meetings
with the community to understand what they wanted and how to do it
safely, so that you don't have to worry about negative effects,
including social impacts of jealousy about different deployments.
devices come with agreements to participate, so it isn't a device
for-free. highly participatory model.
Leandro: important findings from ict4d research community to incorporate
assessment in the perspective of replication.
Kevin: re chat q (Adrian Farrel), project is about adoption by
individuals, not just leaders. Sometimes community tends to go to
leaders, but n50 insists on getting to children, families, individuals
in the community.
Leandro: sharing power as a way of helping everyone in the community,
but an ongoing challenge.
Jane: +1 for community-based approach. recent work in the US. people can
often "parachute in", which doesn't create sustainability. planning
grants for small networks, bringing in a variety of people in the
community is essential, to avoid single points of failure, both in the
tech and in the community. how to assess "community readiness"? need
different tracks of learning simultaneously.
Kevin: experience working with tribal communities in the US. one of the
first questions is 'who are you and how long will you be here?'
sustainability is a necessary trust-building operation. project
curation: community readiness, partner readiness. lots more to talk
about (on website or later).
Rajeev: technical solutions of integration with telcos and deep edge
caching might be useful start for technical standards or best practices
that could become an RFC. where in IETF are community network standards
Leandro: RFCs not a marketing tool, hard to reach non-technical
Rajeev: want to make these projects lower-risk, higher-return so that
they can be scaled.
Leandro: there is a useful gaia RFC in the past, but community networks,
there may be solutions that are well-adapted to those environments. some
work on low- connections, energy-related.
Leandro: information-centric networking; human rights protocol
considerations; decentralized internet -- meetings tomorrow may be
Kevin: completely different solutions for displaced people in/around
Ukraine. millions of people. shipping containers with networking, power
that can be moved and re-connected. service provided around the
container (equipment, not people, inside). can provide wayfinding
services so it's useful for people to find info from local agencies.
Leandro: relevant project:
Kurtis: thanks to presenters and notetakers. Gaia a good place for
giving presentations like these. Jane: join gaia list if you haven't.
could explore/update RFC.
[other presentations not available today]
Subsea data centers
Maxie Reynolds, Subseacloud.com
HolistIX for IXP https://holistix.iijlab.net
Marc Bruyere, U Tokio
See overall discussion above.