Broadband, an Infrastructure Challenge

Last week, the US government started exploring the possi- bility of a new Broadband Bill, and this prompted me to look at the status of our Internet connectivity is. The US mulls moving to a minimum speed of 100 megabits per second (Mbps) to each home and at least a gigabit per second (Gbps) to each university. I am not aware of any university, college or a corporate that is currently using 100 Megabits per second. India is one of the leading countries that develop and work as the backbone of information technology. But while the Internet penetration has grown and speed leaped forward, the way our Internet access has been growing is something to worry about. This week I try to explore what are the reasons I feel that the growth has been slow, and how we can make the growth speedier .

We all agree that the Internet connectivity to the masses, commonly known as broadband—defined as always-on-Internet—delivered to people has helped us grow the economy, our personal businesses as well as increase the reach and availability of information. Some of us will argue that there is still a digital divide and some will take the stand that though Internet tariffs are attractive, the speeds are not.

Internet was made available to the masses on August 15, 1995, through the Gateway Internet Access Service (GIAS) of VSNL. Back then, you could get a 14.4 Kbps going up to about 28.8 Kbps of Internet bandwidth, starting with a limited level of access shell account to the new age TCP/IP socket accounts. Then, you paid roughly about Rs 15,000 per annum for a 500-hour connection.

Zoom in to 1999 when telecom operators started realising the 56 Kbps that the dial-up Internet could reach was limiting their growth. New private players emerged with the first set of broadband services, at 64 Kbps unlimited, but you had to shell out anywhere between Rs 3,000 and Rs 6,000 a month depending on if you were a corporate or a home user; this is where the trouble started.

The Internet does not distinguish between a corporate or a home user; both use the Net to download content and to send information to other networks. But yes, the corporates were assumed to be heavy users while the home users were thought to be light. I could see why the price arbitrage was required back in those days, as roughly 97% of the servers that had content and email were outside India. Data needed to travel on international private leased circuits which would cost an ISP or telecom provider similar to an international call, making bandwidth expensive.

The sudden explosion in the dotcom space, a lot of development of applications and growth of service providers spun off entities that would host the servers in India. My company was one of them starting back in 1996, but we did not see content move to India until as late as 2004 when suddenly most of the large content providers started looking at servers in India to offer higher speed connectivity and a better experience to their website visitors. Also, in keeping with trends, the broadband kept redefining itself to a minimum of 128 Kbps. In 2005, the government of India mandated that for a service to be called broadband, it had to have at least 256 Kbps of interconnect speed between the customer and the point of presence. Another big mistake here.

The 256 Kbps of speed was defined as the capacity of the line between your house or office to the telecom/Internet service provider and not the Internet bandwidth available to you. The government suddenly realised that there was need to interconnect all Internet service providers and within themselves; they were not talking to each other. So, the National Internet Exchange of India (NIXI) was set up in 2003 and today, approximately 31 ISPs connect and exchange information at NIXI, which sadly does not go above 8 gigabits per second on any day. While NIXI backbone has been built to support up to 100 gigabits per second of traffic and stays underutilised, more and more bandwidth needs to be pushed to exchange points.

Most of the urban homes have broadband now, those who don’t still think it is too expensive or because they don’t see a need for it. Cheaper broadband will enable more and more people to connect to it, while faster broadband will enable doctors to monitor paitients remotely over the Internet (What we saw in 3 Idiots is just a glimpse of things to come—people are already using video conference to talk to each other, but the cost is still too high) and students take classes online. High speed internet means high quality video.

Companies such as BSNL and Airtel are at the high-speed frontier. Airtel has announced that the minimum speed of its broadband will be 512 Kbps. BSNL has a 24 Mbps plan, though it may not be easily available. But it is interesting to see a state-run operator coming out with the fastest possible broadband plan aimed at home users.

The ISPs need to drop the differentiation between a home user and a business user. The differentiation could be on the sharing ratios of services: business users end up buying Internet bandwidth as leased circuits, and pay as little as Rs 1.8 lakh per annum for a 2 Mbps link, i.e, Rs 15,000 per month for 2 Mbps, while a home user may get a 2 Mbps unlimited plan for as little as Rs 4,000 per month. It is the same bandwidth, but with different content ratios; while the corporate user will be able to peak up to 2 Mbps at all times, a home user may or may not peak depending on the loads.

The price war is taking place, new ISPs such as Tikona are changing the landscape and older players are entering new territories. But the unfair use of the term ‘fair usage policy’ needs to be looked at seriously. So, unlimited connection means there is no limit, but a *on the ‘unlimited’ signifies that there is a rider; people have billing hassles with the largest ISPs and tend to choose smaller plans, or stick to plans of limited speeds. The chicken-and-egg question of the content or the speed first needs to be answered at a time international connect prices are falling, more content providers looking at India and the country generating its own content. The content and the bandwidth seem to be merging and plans need to be drawn up for truly unlimited downloads.

Where do we stand today? The government is very supportive and ISPs have started talking to each other, but the premise that most of the content still lies outside India is not true. The other premise that international bandwidth is expensive also does not hold true any more. The world’s leading connectivity companies such as Teleglobe Network, i2i Networks and Flag are owned partially or fully by leading Indian telcos such as Tata Communications (formerly known as VSNL), Bharti Airtel and Reliance, respectively.

The premise that content is out of India is changing at a high speed. Though telecom operators need to focus on interconnectivity within each other and setting up peering/exchange points, they are still trying to outsell each other in terms of bandwidth. An archaic law such as the one stipulating content providers will not be allowed to peer with NIXI creates its own headaches. Today, a company such as Google has to come in and set up its servers at all three major telecom providers (Bharti Airtel, Tata Communciations and Reliance) while they could easily plug into the NIXI and be available to all Tier-1, Tier-2, and Tier-3 ISPs. We need to review the opportunities in the infrastructure segment and enable the growth of the Internet.

My request to the government would be to expand speeds in multiples of megabits per second, and let private exchange points come into the picture. It is in the interest of Internet service providers to connect with each other. I agree that they are competing but interconnects at independent levels, similar to what happens on the telephone links, would help the market mature and take a different shape. The definition of broadband if expanded to minimum 2 megabits per second will help increase penetration and expand use of the Internet beyond the casual use of checking email, updating Facebook status, watching live interactive videos, being part of global conferences and creating more and more content.

Broadband is an infrastructure challenge. It has been a long time that it was a demand and supply game. From now, the goal should be getting more people online and the price and speed mixture correct. We have come a long way and the growth has been amazing, but going forward, we need to enable leaps in megabits per second, and not kilobits per second. Rural areas lag not because there is a digital divide or there is no need of broadband there. It’s just that players have not reached this market as yet. The government has announced schemes, but there is a need to have content. Your average Web hosting company still wants to put up its servers outside the country as it is cheaper. But this flow of servers outside not only moves our content outside, but also precious dollars other companies earn. The more content is local and the more content is generated, the more impetus will it generate.

The Above article appeared in the Financial Express, on Thursday 25th March 2010


  1. Your best bet is to get a wireless modem reutor such as a Netgear DG834. You can plug into it for your broadband with an Ethernet cable or use its wireless function to connect your laptop.With AOL (yuk) any reutor not supplied by AOL needs a little bit of extra setting up to get the best speed (MTU setting needs adjusting to 1400 but this is easy, just google it)