India's Electronic Literacy - Contention Rates

I came across an interesting article today over internet speeds in India. There are two conflicting things which decide the speed which I get - bandwidth decided by the provider and the contention ratio - number of subscribers sharing the bandwidth at the same time. In layman's terms, paraphrasing from the same article from Times of India, Say, 10 customers are sold 2 Mbps connections, which means 20 Mbps should be available. But the company may provide only 5 Mbps for these 10 customers, assuming not all customers will be using their internet connections heavily at the same time. Therefore, the guaranteed internet speed is not 2 Mbps, but just 500 kbps.
5 customers using internet at the same time need 50 Mbps line. But, what is provided is only 5 Mbps which means that the contention ratio is 2.5. The company assumes that no more than 2.5 customers use the line at the same time. This is called contention ratio. But, if all the users are logged in at the same time, we are looking at 500 Kbps as against the promised 2 Mbps. Only thing is, in reality, we are looking at may be 10,000 or 20,000 customers as against the ten used as an example.
On a grander scale, it means that unless the lines are undersubscribed or if it is a very slack period like 2 in the night, it is assured that I will never get the speed which I am promised. If I subscribe to a 10 Mbps line and if I get 2 Mbps, I can survive. But, if I subscribe to a 1 Mbps line and get 250 Kbps(or 64 KBPS), life is going to be a hell. This is a reality which we are forced to face day in and day out and coupled with atrocious signal availability and services, we are staring at an even grimmer picture.
Now, the real questions are these.
1. Who decides the contention ratio for an internet provider?
2. Is there a maximum contention ratio allowed? As a converse, is there a minimum speed an individual should get?
3. Going by the fact that cost of usage of 1 MB of data as a percentage of country's per capita income is highest in India when compared to the top ten economies of the world, what is being done to increase the minimum available speed for a customer?
On a grander scale, though India is fast catching up and this data is slightly old, the below table gives an insight as to what India's position is, vis-a-vis the top ten economies of the world.

Rank	Country/Region	Nominal GDP (100 Billion US$)	Cost of Internet ($ per Mbps)	% Per Capita * 1000
1	United States	174.19	0.86	18.90%
2	China	103.8	4.57	722.62%
3	Japan	46.16	0.02	0.66%
4	Germany	38.6	0.27	6.81%
5	Britain	29.45	0.92	24.18%
6	France	28.47	0.27	7.27%
7	Brazil	23.53	0.55	56.88%
8	Italy	21.48	0.54	18.09%
9	India	20.5	0.97	715.43%
10	Russia	18.57	0.24	22.28%

It is interesting to note that based on what you earn, internet is 1000 times cheaper in Japan or 100 times cheaper in France as compared to India. Though this is offset by the fact that internet penetration in India is only 15% drastically cutting the cost, the issue still remains that internet is filthy costly in India. And arising out of the demand, it is but natural that contention ratios are high/not followed since the rate of increase in availability of spectrum is way less than the number of new electronic literates. 

To address this inconvenience which is going to hit India big time in the next few years, the government should give a set of guidelines as to how are we going to manage the increasing subscriber base. ISPs should also do their homework to make it more viable.

1. What is the expected internet penetration in India after 10 years?

2. What is the spectrum availability at the end of the specified time?

3. Assuming an average internet speed of 5 Mbps, what will be the contention ratio for India as a whole? What should be the maximum permissible deviation from the average? 

4. The dashboard provided to the user should detail the current contention rate vis-a-vis planned contention rate. This single tool will kill the errants.

These numbers should be calculated for India as a whole and for the target clientele - urban, rural, remote and business.

This being said, there should be a roadmap to

1. Increase the spectrum availability by the end of the time

2. Investment in alternate forms of network provision - satellite communication, fibre optics(the work is not completely done), subscription based community hotspots etc

3. Investment in technology and support of entrepreneurship for huge scale up of existing provisions like a single hotspot covering, say, an area of 500 sq km. This will drastically reduce the cost of deployment in remote locations