Position Papers/WSFII

From Www.india-gii.org

Note on WSFII discussion meeting at the Ministry of Information Technology

Members of the World Summits for Free Information Infrastructures attended an informal meeting with senior officers of the Ministry for Information Technology and guests from industry, academia and the NGO sector on October 19, 2006. WSFII members who attended thanked the Ministry for its hospitality, and for its openness in discussing contentious issues relating to broadband networking of villages.

The annexure contains a brief note on each presentation made at the workshop. Points for thought and action emanating from the presentations and discussion include:

Points for thought and action emanating from the discussion include:

  1. A delicensing of the spectrum for 802.11a would significantly reduce the cost of long-distance connections from the village to the nearest optic fiber and increase available bandwidth. This technology effectively carries upto 50 megabits per second in both directions, as demonstrated by deployments over several years around the world.
  2. The current policy for delicensed 802.11b/g spectrum puts a limit on the power output after including the antenna gain. This penalizes those who use highly directional antennas, which produce less interference for other users. The level of power in WiFi communications is so low, that the maximum power specified should be independent of antenna gain.
  3. While we think it is a good idea not to waste the copper lines going to some villages, 512 kilobits per second is very little bandwidth for an entire village. As soon as the network becomes successful and starts to attract new users, it will slow to a crawl. Our suggestion would be to interconnect these villages additionally via a wireless mesh network. This would allow other villages to utilize the bandwidth that one was not using, provide significant bandwidth to villages to communicate with each other free of cost, and spur local capacity building and entrepreneurship. It would also provide redundancy, so that if the roof of the telephone exchange falls during an earthquake, a backup means of telecommunication is available.
  4. The manner of spectrum regulation is being critically reexamined all over the world. Auctioning of spectrum puts unbearable burden on any new technology: it may, for instance, have dealt a fatal blow to 3G before it was born. Imagine how high WiMax spectrum would be bid, in open auction. That money would have to be extracted from the pockets of consumers, whose numbers therefore would be small. Also, the current system requires the government to play technology god, deciding, via spectrum allocation, which technology gets a chance in the market, which doesn't. The best way to handle spectrum is to learn from the spectacular success of WiFi, and delicense large sections for shared use, in which the government only sets the parameters for fair usage and sharing. Particular attention needs to be paid to spectrum in the lower frequencies, which allow communication without line of sight, thus reducing the height and cost of masts.
  5. With regard to choice of technology for broadband wireless interconnectivity of villages, we feel that the following criteria are important. In each, we compare WiFi and (non-Line of Sight) WiMax, the technologies that currently come into consideration for this application.
        1. Throughput: Appropriate 802.11a hardware provides as good if not better throughput than currently available WiMax products
        2. Price: A study comparing the cost of using WiFi to network a specific area versus using WiMax to do the same, which was conducted towards the end of 2005, found that the WiMax gear was 20 times as expensive as the WiFi. Unlike WiFi, WiMax hardware must use different frequencies in different countries, thus reducing standardization and increasing cost.
        3. Spectrum availability: It would help, if the Ministry would clarify to us, what the policy of the Indian government is, with regard to WiMax spectrum. Which slices of spectrum are likely to be freed up for it, whether they would be delicensed or auctioned. Until there is clarity in WiMax spectrum, there is no point in planning any WiMax deployment.
        4. Rate of Technological Progress: There is a lot of research taking place worldwide in WiFi-based technology. An example of this is the work being done in mesh networking protocols, such as OLSR and BATMAN. We do not see such a high rate of progress in developments in the WiMax space.
        5.  Support: The actual agencies deploying the communications hardware will be confused about technology choice. How is the Ministry addressing this, and could we help? As of now, support for WiMax is expensive and available from relatively few sources. Against this, expertise in and support for WiFi-based networking is ubiquitous, and often available free, or at little cost. 

To summarize, we would urge that the government delicense outdoor use of 802.11a, stop penalizing high-gain directional antennas, provide redundancy via mesh networks, delicense spectrum rather than auction it, and examine carefully the merits of WiFi versus WiMax before committing significant resources to the hardware.

Annexure: Summary of Individual Presentations

Malcolm Matson, from The OPLAN Foundation, who has been advising the World Bank on the local /open access infrastructure/ business model and its importance for the developing, posed the important question, "how does India get the */greatest/* benefit from the deployment of the disruptive digital technologies (silicon chip + fibre + spread-spectrum wireless) - letting one industry (the current telecoms sector) extract the value for itself /OR, /adopting a new business model supported by public policy which would leave the value within society and the economy in general. Put another way - is the best value for India from radio spectrum to be gained by selling it to licensed operators (and generating 'fees' for the Government) /or /making more spectrum license exempt and thereby hugely benefiting local economies? Malcolm Matson was emphatic in saying that the latter was preferable and that from his perspective, looking at what is happening around the world, those nations that adopt an open access approach and promote license exempt spectrum will enjoy faster economic and social development than those that hold back in order simply to support their existing telecoms sector.

Elektra said: Internet access penetration should be distributed by a three tier approach:

1. Use of fiber line wherever possible.

2. Wireless long shots to distribute high-speed access wherever fiber is not feasible. Delicensing of 802.11a for outdoor use would allow to have relatively cheap wireless backhaul links at high speed.

The benefits are:

• Higher frequencies compared to 2.4 GHz WiFi have a significantly smaller fresnel zone, thus the height of towers could be smaller and

therefore the towers are cheaper.

• More available non-overlapping channels. 2.4 GHz WiFi has 3 non-overlapping channels, while 5.X GHz has typically 8 non-overlapping

channels in the frequency range of 5.725 to 5.850 GHz (depending on regulations).

3. Use of inexpensive mesh technology to distribute the network locally in rural areas. Cheap of-the-shelf hardware could be used together with open-source operating systems and mesh software to spread ubiquitous network connectivity at low cost.

Beside this it was mentioned that it is better to limit the output power of WiFi transmitters or WiFi amplifiers rather than calculated effective isotropic related power values (E.I.R.P.) which limits the gain of antennas. Strong directional antennas create by far less interference than amplifiers.

Bjarke Nielsen reported on the DjurslandS.net in rural Denmark, in an area where the market fails to provide Internet access in most of the countryside.

This not-for-profit outdoor wireless community network is run by local volunteers - in their own interest. It gives 4 megabit broadband access,

both downstrean and upstream, to each of about 5000 households, in an area of 60*50 kilometers. Each household's contribution over a 4 year period is

surprisingly only 1/3 of the average market prices in the cities. He stated that the high quality cheap network model from Djursland might be

applied and moderated for rural India in the Mission 2007 program, making the poor rural areas benefit the most as soon as possible from ICT

familarity. He ended his contribution by giving a warning: Don't waste big money on expensive WiMax solutions, as research has shown that it will not

provide a better solution for the users than the very cheap WiFi solution on Djursland.