Paper for the "ITS 12th European Regional Conference"- Dublin,Ireland; September 2-3, 2001
International Telecommunications Society     www.itsEurope.org

NETWEAVING RURAL VILLAGES:
INTRODUCING THE WEAVE


Ir. Jaap van Till,  Stratix Consulting Group B.V.
P.O. Box 75554, 1118 ZP Amsterdam Airport, The Netherlands
Phone +31 20 44 66 555, fax +31 20 44 66 560,  jaap.vantill@stratix.nl


Abstract
Teledensity in remote areas of the world is very low and it is not expected that that situation will change very soon, because present telecom investments do tend to go to large cities and regional population centres first. Therefore many people in rural villages will not have the opportunity to improve the quality of their life and their local community by using telecom for social and economic transactions, among each other and long distance. The presented innovative telecom system is defined specifically to help "wire" local communities in remote and rural areas, by way of wireless voice- and e-mail messaging. The peer-to-peer routing and self relaying nature of the proposed solar-powered grid of devices, without the need for central infrastructures, gives this new telecom system the potential to bring low-cost telecom and internet functions to billions of people on this planet. And this new local Weave grid can interconnect to and interwork with the present telephony and Internet networks, creating traffic and new services there too. Possible worldwide impact on teledensity and some of its policy consequences are presented.

1.  INTRODUCTION AND STATEMENT OF THE PROBLEM
Often the problem of the "Digital Divide" in the world is mentioned as the huge difference in teledensity and computer density between rich and poor nations.

A closer look reveals that the ICT density difference within countries,  between cities and rural villages & areas is even more striking. This  a problem since it is recognized that an important factor in social, cultural  and economic prosperity, and thus the quality of life, is the ability of people of cooperate, transact and communicate locally and long distance.  And if life in rural areas and remote villages is not feasible people will continue to move to large cities in search for work, food and sustaining systems. In many developing countries plans have been announced to bring "at least one telephone to every remote village" in the coming decades. This will require enormous efforts and government investments and even then it will not help many families who live in forests or in the mountains or who move around with their cattle to reach other family members or villagers many miles away. And commercial investors do not see these remote needs as their first priority either. The profitability, and therefore investment ranking, from high to low, in telecom infrastructure is: Not surprisingly the teledensity of rural areas in developing countries is hopelessly inadequate, and will stay that way in the foreseeable future if we stick to present telecom technology systems only. This has two big disadvantages for the population. It is very difficult for villagers to reach emergency-services, to reach for knowledge to solve problems or to conduct trade or organise logistics when they have to walk for hours or even days to reach a telephone to contact specialists or markets concentrated in the cities. And in the other direction, those villagers can not be reached by the specialists and markets from the cities. The other obstacle is that communication between local villagers in their own communities, sometimes living or travelling far apart is difficult too. Social, cultural and economic transaction costs in time and money stay too high for local cooperation to flourish and communities to prosper.

This paper analyses the situation and proposes a new telecom solution for hundreds of millions of villagers, which is feasible and can be made low cost with a clever combination of recently developed available technologies. An earlier version of the idea presented here was published on the Web, called "the TELLET proposal"  /1/.

2.  The present situation in voice telephony and Internet access.
Besides the networks for radio- and TV-broadcasting two other worldsize  infrastructures are being rolled out over the word.  Fixed line- and  mobile voice telephony now has in total about 1100 million and 800 million  of subscribers respectively, with rapid growth in sales of GSM mobile phones  all over the world, so mobile phones in some countries even exceed the number  of fixed line phone connections. Teledensity (fixed + mobile telephones /  100 inhabitants) is spread very unequally between 103 lines /100 inhabitants  in some rich countries to 0.01 % in remote areas of poor countries. Internet  access now has a penetration of about 300 million PC's connected, also with  a fast growth rate all over the world. /2/3/

Obstacles to further penetration are:
Consequences are that both voice telephony and Internet use are relatively expensive and concentrated mainly in the larger cities for people who can afford ICT and can use it in government or creating more wealth in those cities. Indeed the economies and populations of most countries get interconnected and interlinked more globally every day, except for the remote areas of those countries which are left behind. Young people and craftsmen leave the villages for education or better jobs in the larger centres of population.

One of the largest problems we have to face is how we can try to keep remote villages feasible to live in. The main trend is that villagers all over the world travel and try their luck in the already overcrowded large and even superlarge cities of our world.

There are a large number of national, ISOC, UN and ITU co-ordinated efforts  put into the further spreading of telephony and Internet access, but   these will take decades to reach remote areas and villages in deserts and  jungles of Asia, Africa and South America. The Republic of China has for instance a ten-year plan to put at least one phone in every village of their western areas.

Conclusion: billions of people will have very limited or even NO access to telephony or Internet for the next ten to twenty years, unless we do something about it.

The challenge therefore is to try to interconnect, to weave the people locally together first and then connect this "Local Village" to other villages and the Global Village. This will be useful to support messages, transactions and family ties for local social, cultural and economic cohesion. This should not be considered as "anti-globalisation", but a plea to urgently fill the vacuum at the lower end of the spectrum too. Not directly aimed at the benefit of the affluent nations and their huge telecom backbones but at the well-being of the rural communities, which are not well-connected to us at all yet.

So the problem stated here is: can we construct very low cost telecom devices which can be introduced and spread out literally "bottom up" from the desert into the villages into the towns? And can the villagers reach out with these devices and connect into the networks of the rest of the world too?

3. ANALYSIS AND PROPOSED SOLUTION
If we look really closely at voice telephony at present,  voice-mail messages or answering machines functions can cover a large part of the functionality of person-to-person conversations. Big advantage of 'messaging' is that we do not have to be present or have to interrupt other activities when such conversations are not conducted in real time. Just imagine a telephone system, which runs with spoken messages only.  Such a system could do without end-to-end connections through 'cleared' circuits and switches, which are very costly and dimensioned in terms of transmission delay and switching capacity. So in this case the telecom infrastructure investments would be very much reduced since only store and forward of non-urgent voicemail is required.  In such a NON-realtime network we could very easily add SMS messaging and E-mail messaging for those users who are literate. E-mail is the most used and most important component of Internet use in the world, and it does not require high capacity links either if we leave out real-time WWW.  If we can implement such local rural message-only networks the users can exchange voice messages and SMS or e-mail messages not only locally, but world wide too with the present telephony and Internet subscribers!! A remote message-only network could be viewed therefore as a access ring around the present telecom networks.

If we break away from the usual assumption that we need complete functionality  of telephony and Internet, we can  do without the very costly network  infrastructure and PC components and still get the most important partial  functionalities (voice messages and e-mail communication) at a fraction of  the costs at outskirts of the remote areas as a stepping stone towards later  full functionality worldwide. This message-only network is the essential idea of the "TELLET proposal", mentioned before.
To implement such a network the proposal is to introduce very low cost wireless devices with only partial voice functionality (no direct conversations, but stored voice messages) and partial Internet functionality (no PC, no direct PC-WebPages interaction, but only SMS, E-mail and possibly e-mail enabled data access). Since the network requirements for non-realtime traffic are very low, we propose to implement the function of the telecom infrastructure into the proposed wireless devices themselves!
This means implementation of peer-to-peer switching and -routing of messages over wireless links between the portable devices.  No further telecom infrastructure investment and maintenance locally would be necessary.

This idea of a 'network with end-user terminals only and no infrastructure' is not as theoretical a it may seem. In fact such an IP communication system is supposed to be in use for instance by the U.S. Army mounted on vehicles when they drive into a battlefield, like in the Iraki desert. It is obvious that they have to carry their own telecom infrastructure with them. So why not develop a similar yet portable structure to hunters and gatherers in the Gobi desert or in the mountains of Malawi?

4. EXAMPLE OF IMPLEMENTATION AND SYSTEM OUTLINE
Proposed in this paper is to give or sell to every adult in remote villages  a personal addressable device which is solar powered or handpowered. The individual devices can be  the size of for example Palm organizer handsets or GSM handsets and are linked by digital radio communication (peer-to-peer packet radio like BlueTooth or Wi-Fi =WLAN IEEE 802.11b).
Most of  the components and subsystems of this proposed network are already available. For instance the company Rooftop, now part of Nokia, sells routers which can be linked into grids for packet-radio IP wireless access to Internet. Systems sold by Aironet, now part of Cisco, have similar functionality.

Essential of the proposed network is that the devices use each other as store and forward relays/routers (peer-to-peer switching and peer-to-peer routing. The more devices in an area the better the parallel total transmission capacity of the network is, without the need for other telecom infrastructure. In fact the devices are their own infrastructure!  And this grid grows simply by adding more devices, thus automatically absorbing the extra traffic of the added devices. By overlap between village networks this "Weave" can spread bottom-up over the globe in a few years.

Such remote local wireless peer-to-peer message grids can be interconnected in many places to the national telephone and Internet networks, creating paid traffic and services there. So its appearance would not pose a threat to the present network operators and carriers but would on the contrary boost their business. Their subscribers can get access to hundreds of millions of new entrants to telecom. New services can be implemented like for instance e-mail- and voice-mail language translation which may be feasible since exchange of these messages is non-realtime and therefore allows delays for processing.

Main function of the device of the Tellet Project local village Weave network would be a local "intercom" between family members which can be miles apart in the field. By easily pre-selected choice on the device one or several family members can be sent a voice message. Or after "dialling": selecting the number of the person you want to reach, you can enter a speech message into the device for transmission, for instance after you have played it back to hear if it is recorded correctly.
If  the voice signal is coded by DSP chip this gives a IP stream of 8 Kb/s. A message for instance of 50 seconds would be stored on the receiving or relaying devices as 40 K Bytes.
At the portable device of the destination an optical signal can signify that a personal voice-message is waiting, like on an answering machine.  Locally delay can be short so message reception and answering could be quick enough to approach a conversation. But also the incoming messages can be played several times or collected and stored for later reply.
And SMS plus E-mail can be added to the functions. One very interesting possibility of non-realtime e-mail messages is the future addition of language translation to this network, making its scope world-wide indeed. And it may be possible to connect PCs to this grid later to use the massively parallel paths of the Weave for IP-transmission to connect with the World Wide Web.
The described new telecom system 'for the rest of the World' has most of the characteristics of a "disruptive technology", as described by /4/ : new user groups, simple to use functions  and low thresholds to implement, but less functionality at first, and no demand for it at first.

5. FEASIBILITY QUESTIONS
(1)    Do members of small remote villages want such a personal "halfway" telecom system that would allow them to exchange voice and text messages among each-other and long distance, or would they rather wait for a single phone box in the village centre?
An important question for this new disruptive technology proposal is:  Is 'half a glass'  half full or half empty?? Are the remote villagers interested in such partial functionality: no direct voice conversations, no direct WWW access?  Is the large scale introduction and use of PARTIAL Phone/Internet functions in arid and remote areas welcome and useful, and much better than 'nothing, no connection' or is such intermediary level perceived as unacceptable, rubbish, below grade ? Or put more simply: Can and will the villagers pay for these 'wireless local intercom' devices, would the devices present value for them ??

(2)    Can we foresee, imagine and trigger a number of completely new "value chains" based on the rollout of the proposed Weave network?

(3)    Can we get funding to publish and promote the spreading of this idea, and encourage universities and regulatory organisations to define standards and interfaces; and would that encourage companies to start R&D projects to make prototypes and do field trials? Or would publication spoil the chances for companies to establish patents? Or put more directly: is there incentive to invest in the development, production and sale of the devices?

We assume that the answers to these questions can only be found by starting small scale field trials in remote areas with prototype devices. Start a race. Like when the first PC's where constructed and tried. And at the same time starting a rigorous standardisation process. Like the 'Groupe Speciale Mobile" did, that defined the GSM specs and network Schnittstellen.

6. CONTEXT AND POLICY IMPLICATIONS
Content is not king, but short or long distance communication between people is /5/6/. The value of such communication is firmly perceived and its participants in dialogues do pay substantially for telecom and internet services which support them. Whether this will be the case for access to information or consumption of broadcasts remains to be seen. Messages, either in voice- or in E-mail form, can be considered most important and valuable to the participants. Therefore we suggest the following new definitions to measure certain functions of "Teledensity", until now defined as the number of voice telephony lines/terminals / 1000 inhabitants in specific geographic areas, like nation states:

(a.) Digital Teledensity = the number of persons addressable with voice-mail or e-mail / 1000
     inhabitants in a specific area. The unit for [T] is dimensionless.
(b.) Digital Access Capacity = the sum of the peak access capacity of the the end-user devices in an
     area expressed in [ Mbit/sec / km^2]= [R], after Hendrik Rood.  For instance PC's on Ethernet
     would count for 10 Mb/s ; GSM handsets for 32 Kb/s.
(c.) Digital Network Power =  the total packet per second power of all the routers in a specific area.
     The power of these routers is expressed in IP-packets/second or pps. The unit for 'Digital
     Network Power' is [pps/km^2] = [B]  after Paul Baran who first defined packet-switching.

It is a common phenomenon that if we gather statistics of items and sort them according to size or frequency of occurance we see a "rank order graph" which is very unequal. Like the size of cities in the world of which only a few are very big, or the frequency of characters in a book. E and A are very popular and for instance Q and K are occur not very often. This is what most people know as the '80/20 Law', which is used often in Quality Control (Pareto charts) and in coding.  In popular terms we can say that success is spread very unevenly. Such sorted rank order charts roughly follow and inverse size distribution: the size of item N is related to 1/N. This relation was empirically found by the linguist Zipf and is called the 'Zipf rank order Law', which was proven by /7/ which is based on the fractal scaling characteristics of "networks" and other types of clusters /8/.

If we were to count and chart T, R and B according to rank order for a number  of areas in (and within) countries we would also see very skewed (very unequal  teledensity, access and networkpower) graphs which would broadly relate to  1/N if N is the rank. While areas can dynamically change places in such sorted  graphs the general pattern seems to stay rather stable, as can be seen in  /3/.  Which means that by for instance cross-subsidies the more or less  succesful (with lower teledensity T and R, B ) areas may change some places  but the inequality between them is not really lowered.

There is however an other solution for the very low wired-ness /9/ of the areas in the long low tails of these graphs. We can together try to lift the total graphs up ! That is essentially the effect that the implementation of the proposed devices and grids in this paper may have. T, R an B of ALL areas will go up, if we were to succeed in adding hundreds of millions of simple communicating devices in the rural capillaries of this world.  The proposed idea is significant, feasible and can have a huge impact on the developing countries and is recommend for further study, since like in the human body the health of the whole 'web of life' depends very much on the health of the micro-communication veins of this planet.

References

/1/ [van Till 2000] The TELLET Project Proposal, version 2.3, July 23 , 2000,   
     http://huizen.dds.nl/~vantill/divide.html
/2/ [OESO 1998]. Internet Infrastructure Indicators, report DIST/ICCP/TISP(98)/Final. OESO: 
     Working Party on Telecommunications and Information Services Policies
/3/ [Siemens, 2000] "2000 International Telecom Statistics", Yearbook Dec. 1999; Siemens AG,
     Munich, 2000
/4/ [Christensen,1997] The Innovator's Dilemma, Harvard Business School Press, USA
/5/ [van Till, Op Hey, 1987] 'Prosumer Networks' -Broadcasting Model vs. Communication Model-
    IFIP TC 9 Conference on Social Implications of 'Home Interactive Telematics' (HIT), in 
    Amsterdam, The Netherlands, June 24-27, 1987, Concerning Home Telematics, F. van Rijn and
    R. Williams (eds.); pp. 361- 367; Elsevier Science Publishers B.V. (North-Holland) © IFIP,
    1988. Also on: http://huizen.dds.nl/~vantill/prosum.html
/6/ [Odlyzko, 2001] 'Content is Not King', Andrew Odlyzko, AT&T Labs,  online magazine First  
    Monday, vol. 2, no. 6, Feb. 2001. URL:  http://firstmonday.org/issues/issue6_2/odlyzko/ 
    or at his home page: http://www.research.att.com/~amo  The Economist mentioned it in their   
    Dec. 16, 2000 edition
/7/ [Mandelbrot, 1997]  Fractals and Scaling in Finance - Discontinuity, Concentration, Risk -
    1997 Springer-Verlag
/8/ [van Till, 1997]  'Fractanomics' - The Issue of Scale in the Network Economy, November 12,
    1997.  http://huizen.dds.nl/~vantill/fractanomics.html
/9/ [Jhunjhunwala, 2001] 'Telecom Technologies for Developing Countries';
    Prof. Ashok Jhunjhunwala, IIT in Madras, India Conference: Digital Infrastructures,
       Vanguard Technology Transfer Institute, Dublin,  Juli  18-19, 2001