From A2K Wiki

Notes on panel

Introduction by Shyam

• Two ?s:
  • How can ICTs be used to further A2K?
  • Does current ICT discourse need to be reoriented?
• Moderator is Prof. David Hakken

David Hakken

• This panel has been beautifully set up by the earlier panel on network neutrality

Catherine Nyaki Adeya, ICT consultant

• Connection b/t wireless ICTs and A2K is huge topic
• Going to focus on a few examples that I know
• Start w/ two photos of Nairobi - classic case of digital divide
  • One side has radio, TV, telefone lines, PCs, Internet
  • Other side has very little but is where wireless is groing fastest
• Prof. Wilson's book ('The Information Revolution and Developing Countries') suggests that certain things are required:
  • Physical access
    • Lots of focus on this in the West
  • Financial access
    • Capacity to sustain payments for commercial and subsidised services
    • Cheap mobile phones in Kenya cost one-month salery of rural school teacher ($50)
    • Costs $0.10 for 5 minutes of slow internet access
  • Cognitive access
    • Example of a women who works for an NGO on female genital mutilation who went into a cyber-cafe and found sites about her own work by searching on the web
  • Design access
  • Content access
    • Access to relevant sites in the user's own language
  • Production access
  • Institutional access
  • Political access
    • Greater chance of sustained and reliable knowledge access if people have political access
• Most documented form of wireless use in Africa is the mobile phone
  • About 104 mobile networks serving 52 M people in Africa
  • For most new subscribers, mobile is first and only telephone
  • Need more research on access via mobile - many phones are shared by many users so numbers are under-representative
• Lots of interesting experiences in sub-Saharran Africa that need to be documented
  • Ex: 'Telephone of the People' (simia gamen <sp?>) in Kenya

Amos Anyimadu, Political Scientist from Univ. of Ghana

• I have no doubt that there is a very important revolution happening in Ghana, but the gov't will be the last to know
  • I spent 4 hours recently speaking w/ fishers about their use of mobile phones
• Connectivity does not insure community
  • What we need is a community of knowledge and for that we need to build social capital
  • ICTs can play a role in this but they are not enough
    • One important contribution that ICTs can make is to enhance trust in our societies
    • What I found w/ the fishermen is that people exhibit lots of trust in their informal relationships
    • When it comes to institutions there is a clear lack of trust, even though there is generally trust in government
• Important to distinguish b/t different knowledge societies
  • What we are striving for in Ghana is very different from what Northern countries are looking for
  • The South is not on same trajectory as the North
• Distinguish b/t knowledge and information
  • eAfrica vision - important distinction, how to move to a 'wise society'
• Push to wireless in terms of carriage
  • Public telephones taken out b/c they are not being used
  • Instead mobile telephony through wireless (GSM fixed) has taken over
  • Wireless in this context has become the killer application
  • There is enormous hope for mobile telephony in Africa

Rohan Samarjiva, Executive Director, LIRNEasia (Learning Initiatives on Reforms for Network Economics)

• How do poor S. Asians communicate?
  • 49% fixed; 19% mobile; 66% 'public' access
    • Not representative necessarily of the rest of India and Sri Lanka
  • Therefore, wireless is only one component of the solution
    • But becoming more important
• Many in West don't understand the significance of 'backbone' infrastructure in the South
  • Using wireless in the backbone as well as in the access network
  • Much of so-called fixed growth is driven by wireless b/c of CDMA
• Strange case of Indonesia
  • When we think of infrastructure, we think of fixed access networks
    • Copper or coax cable to homes, but wireless only at the ends
  • Reality in Indonesia is quite different
    • Wireless inside connecting ISPs as well as at the end of the network
    • School buys wireless link at $4000/mo and then resails it to local homes and corporations in order to make it affordable
      • The ultimate work around
  • So people are using wirelss in a very peculiar way that is a response to bad institutional governance
    • Un-legal rather than illegal or legal
• Spectrum management is not enough
  • There has to be investment - investment is what connects people
  • What are conditions for investment?
• Conclusion
  • Got to pay attention to institutional considerations
  • Need a decent regulatory environment
    • Even work-arounds in these situations is prohibitively expensive

Rahul Toniga, Carnegie Mellon, Dept. of Engineering & Public Policy

• Wireless and Access (to Knowledge)
  • Do mobiles solve the problem of the digital divide?
• 4 Cs (Computers - devices; Content - relevance is important; Connectivity - ideally broadband; (human) Capacity - literacy, empowerment, etc.
• 4 As (Awareness; Availability; Accessibility - language and literacy; Affordability
  • ICT is a means and not an end
  • Mobile fits many but not all of these
    • What about non-voice content?
    • How affordable is it?
• Are markets working?
  • Yes, but... what about granularity? Could alternatives have done better?
  • Consider India - largest market today, but has very low avg revenue per user ($8)
    • Growth is really in urban areas not so much in rural
• Wirless is important for emerging regions (leapfrog)
  • BUT Wireless requires massive investments
    • Uplinking is still very difficult
  • Reach and speed are inherently in conflict
    • Providors would ratehr provide 'per bit'
  • What is universal coverage?
    • Mobiles are rarely a 'must serve' (neither is broadband)
• Wireless Technology: Hope or Hype?
  • WiMax is classic example of raised expectations
    • "Up to 70 Mbps, 50 kms" - NOT AT SAME TIME
  • Wireless has hidden or implicit end-user cost
    • End user equipment has short shelf life (also batteries)
  • Spectrum is a bottlenet
    • It is a man-made bottleneck
      • E.g., opening up the UHF (~700 MHz band) would improve reach by a factor of ~3-5
• Wireless: Select lessons
  • Wireless cannot stand on its own
    • Needs uplinking
    • Fiber is cheaper than people think
  • Devices matter
    • Mobile vs. general purpase PCs vs. Set-top boxes (PCTV)
  • Technology is improving rapidly
    • Mesh networking
    • MIMO (intelligent antennae)
  • Policy issues are a greater challenge
    • Spectrum
    • Allowed services (e.g., VoWLAN, or even VoIP)
• Fundamental ? - business as usual or leapfrog?
  • Are they designed for 1%, 10% or 100% of population
  • Are they geared toward profit max or welfare max
    • Auctioning spectrum = barriers to entry
  • Fundamental reasons why open access networking may be the best solution
    • Treat core bandwidth as a public utility = everywhere
    • Allow competition for retail services
    • Analagous to roads but a lot cheaper
• Wiring Africa - Proposal: FiberAfrica
  • Cost $1B for the entire continet (excludes end user tech costs)
  • Hybrid of optical fibers and fixed wireless
  • Can be cheaper by harnessing existing infrastructure
  • Revolutionary business model could allow virtually free access to schools and rural community centers
  • Business model is sustainable
    • Public-private partnership
    • Public core, private edge
  • Fiber lasts 30+ years, while electronics amortized in 5-7 years
  • No real conflict w/ competition
  • Can justify "special regulation" for public network
• Conclusion: Wireless will be a key component of solution... but infrastructure is only a small component of the overall ecosystem

Dorothy Gordon, Director-General, Advanced Information Technology Institute (AITI), Gahna

• Why focus on wireless?
  • Africa's share of mobile network worldwide is only 4% and has lowest mobile penetration rates
  • Also, Internet penetration is by far the lowest
  • Only 4% of world telecom investment
  • We haven't really answered this ?
• Gov'ts are looking at things from an economic growth perspective
  • Looking to exploit knowledge economy
  • Not focused on equity issues - trickle-down approach
  • But still a lot of isues w/ cost and effectiveness - we need more research
  • We have higher cost sevice despite lower incomes
• Issues
  • Supply and Demand
  • Fascinating issue - looking at bandwidth as a public good
• How much useful knowledge is there?
  • Very little systematic attention given to local language content
  • Very little understanding of the adaptation of technology that has to take place if communities are going to be able to exploit eventual avaiable bandwidth
• Two polit programs to note:
  • Intel built a WiMax school in Ghana
  • IICD built mesh-network in Ghana

Questions:

• Rohan - nothing new about public funding of infrastructure. Funding is a seperate question that has to do with budgetary priorities (phones vs. water). Once its been built, the question about how to open it up is essentially about creating effective regulatory regimes.
  • Rahul - Private companies won't go into rural areas
• Did Grameen village phone make any difference for the women who participated?
  • [Answer from audience] Absolutely, per-capita income of sellers increased tremendously
  • Dorothy - they are providing a useful service, people can get info they need w/o having to travel
  • Amos - I've done some research on impact of 'information cures' an
  • Catherine - Too much of research has been done on Grameen
• Dorothy - apparently gov't of India zoned the country and invested in those areas that were not competitive enough to attract private industry, and in third zone gov't did it alone. What do we know about this?
  • Rohan - the real issue is that the Indian gov't owned company has done a good job providing service. Gov't money has been used to take fiber out. Gov't is funding incumbants and not neew entrants.
• Rik Panganiban, SSRC - First, in terms of potential unintended market effects - getting rid of middle-men through use of wireless tech; wireless as organizing tool for civil society (People Power and Orange Revolution); finally, role of wireless tech in emergency/disaster scenarios?
  • Rohan - switch networks fail in disaster scenarios, SMS is not optimal. Cell broadcasting on the other hand could be better - its being researched now. Relief and recovery (2nd phase) requires multifacited approach.
  • Amos - I am very skeptical of the getting rid of middle-man approach. What we need now is peer-to-peer content. Fishers use w/in closed, enhanced trust network. Too much money going to top-down approaches.
  • Catherine - Some excellent SMS initiatives in Kenya - using SMS to alert people about health screenings and services. 'Market' SMS in Kenya.
  • Dorothy - Getting market information is just the tip of the iceberg. Its irrelevant to know what the price is when there is only one buyer who shows up to buy.
  • Rahul - differnt things work differnt ways in different places.
• Eddan - in terms of content, user-generated content production seems to have policy space. We see structural similarity of users not being seen as uploaders in many other areas of A2K. Wondering about extent to which participitory user is being taken into consideration in design of these novel networks? Second ? about using phone rather than computer and what the implications are for upload/download as opposed to lateraly communicaiton?
  • Dorothy - commercial companies are developing applications to address upload challenges but we should do more from public sector. Problem is that rural communities often times don't feel that they have anything of value to share, so this needs to be addressed first.
• Caio - Whether it is possible to use mobile phones as nodes in a mesh network? Big bottleneck is 'power' (electricity). If this is possible, how could we convince mobile operators to make phones compatible w/ this purpose? Is there anyway to get the incumbants to play along?
  • Rahul - more fundamental issue is what each technology is best for. The architecture for mobiles today doesn't permit this b/c made to talk to centralized network. If you're building a mesh network, cost of device is $20 so not sure why you need to use the GSM network. Many wifi mesh networks gloss over issues of rights and responsibilities.
  • Amos - mobile companies are making so much money on voice that they are not interested in data.
  • Catherine - just wanted to emphasize the need for more research from a user perspective and more collaborative N-S research. One proposal I really liked was about 'access specialists' and I think this is a great idea.

Papers and resources