By Charles AssisiEvolution is as painful as it is dramatic. It is catalysed by a mutation which is accompanied by both death and metamorphosis. The elimination of the weak is implicit in this process.
There is a catch to it though. In all the chaos, the weak try their hardest to survive. If it means adapting, so be it. That explains human evolution. As for the others, history is witness to what happened to the dodo and the dinosaur.
There is something as real, happening real time, in a very real world -- the Internet. It hitched a free ride to popularity on the back of telephone networks that were meant to carry a different kind of traffic. For engineers managing this traffic, it's been a nightmare.
Over the last twenty odd years, demand for bandwidth has increased one million times; fax transmissions make up fully half of what is considered voice traffic; and in a few years, 99 per cent of all traffic will be carried on Internet Protocol (IP) networks. Over the last couple of years, data traffic has been growing at 85 per cent and in 1998 it outstripped voice transmissions.
What these statistics have done is, upset a whole lot of numbers. These numbers have the potential to destroy existing business practices and throw up new models that may not tolerate yesterday and today. More pertinently, these numbers ought to force governments to do some introspection. And why?
The commoditisation of bandwidth
Quite clearly, to sustain the kind of growth on the Internet, Planet Earth needs an awful lot of bandwidth. Crudely put, bandwidth is a lot like any pipeline. The only difference is that while pipelines carry anything from water to crude, bandwidth transports information. Bandwidth will determine how far and how fast companies can go with real-time video conferencing, cross-continental design collaboration, and the facile manipulation of far-flung databases. Technically, bandwidth is the term for the total payload of data, in terms of volume and speed, that can be sent across a network.
Which is why, there is the $1.7 billion Loral Space, the $445 million Eutelsat, and the $767 million PanAmSat wiring up the skies.
Under the sea, there is Project Oxygen, a 158,000-km fibre optic network with 101 terminal points in 74 countries. Not to forget the SEA-ME-WE, a 17,658 km optical fibre cable system that connects countries from Singapore in the East to France in the West. Next on the cards is SEA-ME-WE 3, which will provide 10 Gbps per fibre connectivity to countries spread across Australia, the Far East, the Gulf and Europe. There are others - FLAG, SAFE, IOCOM, Gulf Cable, etc.
On land, public utility companies like power, gas, the more traditional telecom companies, and young upstarts are doing their best to wean away customers with promises of assured bandwidth. In the US, new players like Qwest, Level 3, and Williams Communications have announced plans to create four million miles of optical fibre.
What is frightening traditional telecom players is that the spate of capacities coming up, threaten their way of life and business. Take Project Oxygen. When it commences operations, the systems will be state-of-the-art and customers will be able to buy just the kind of capacities they need. The pricing systems being instituted are not distance and destination sensitive. Instead, users can send anything they want for as long as they want as long as they don't exceed their bandwidth capacity agreement.
Historically, telecommunication prices have been based on distance rather than the nature of the application. Theoretically, the further a signal has to travel, the higher the switching and transmission costs.
These economic arguments are no longer valid, primarily because fresh capacities are creating an oversupply of raw, or dark fibre. It works like this.
When a national backbone is created, there are typically 48 to 96 fibres. Only about 2-4 are "lit up" or used. The unused capacity, called dark fibre, is up for sale to anyone willing to pay the price. Companies like Arbinet, Band-X and Global Clearing House have already created exchanges that will trade in the dark fibre.
In effect what they have done is create an anonymous marketplace, where the carriers can bring short-term contracts and sell it at whatever price the market is willing to pay. In this market, all bandwidth is created equal and there is no brandname associated with any purchase.
There is another advantage in staying anonymous. They can undercut their own prices to sell excess capacity quickly. Enron wants to create a bandwidth futures market in which buyers can lock in the price of future contracts, allowing sellers and buyers to hedge against the risk of volatile prices.
These exchanges have made traditional long distance carriers like Sprint and MCI WorldCom consider upgrading their own networks. They are also desperately trying to brand their networks. Juxtaposed against what's happening in terms of a commodity market, it may turn out to be a futile exercise.
A McKinsey report argues that traditional players ought to consider getting out of the long haul transport business and abandon all thought of building their own networks. This, for three reasons.
The upstarts have built their networks on fibre optic cable that can support the latest advances in technology. Upgrading older networks will be painfully expensive.
If all of the announced capacity additions fructify, McKinsey estimates an oversupply for the next seven to ten years.
Customer acquisition costs are going to soar and the battle will be fierce. It makes more sense for traditional players to think of how to retain the base they have where the upstarts have nothing to write home about.The bottomline is that transition is going to be far too painful. After operating in a business where margins were healthier than that in the drug trade, commoditisation is not funny. There is a choice though. Let evolutionary dynamics take over.
Coca Cola-nisation of bandwidth
In a recent paper Kenneth Cukier, senior editor and Paris Correspondent for Communications Week International, made a pertinent point. As much as it sounds politically incorrect, fact of the matter is, when it comes to bandwidth, it's a clear case of American colonialism. There is a very tangible reason for this.
The US backbone comprises tier-one Internet Service Providers (ISPs) like MCI WorldCom's UUNet, Sprint, and GTE Internetworking. Between them, these players control a vast majority of the Internet's routing table. This means that traffic on their network can originate or terminate without connecting to any other network.
But for ISPs in other parts of the world, it is almost mandatory to interconnect with one of these tier-one ISPs. It would otherwise be impossible for them to reach either the world's largest number of users or the most popular Internet content which is hosted on the US backbone either by the tier-one or the downstream American ISPs.
Cukier explains these anomalies as a function of recent telecom history. His argument is that today's physical telecom networks were built on assumptions about international call traffic patterns.
For instance, Americans trade with Britons; USA and UK share a common language, and both are among the richest nations in the world. Which is why, the telecom route between the two countries has been lucrative and there is an incentive to create increasing capacities. The costs, after all, can be recouped.
In case of countries that do not possess these characteristics, telecom infrastructure is invariably poor. For instance, there has been no incentive to build a super efficient network between India and Turkey or for that matter between India and Vietnam. But there has always been a clear case for maintaining excellent links with the US because it has been the most popular calling destination.
As mentioned earlier, when the Internet rode to popularity, it was on the back of existing physical networks that carry voice.
All of this has meant that in spite of the fact that the Internet is essentially meant to be a meshed entity, it has not remained one. Instead, the US now forms the nucleus around with all Internet and e-commerce activity exists.
This is disastrous for two reasons. One, domestic companies end up paying US carriers. It means flight of capital in terms of telecommunication fees. And two, in the future, assuming cyber taxes happen, most of these transactions, at least technically, may be taxable in the US. After all the content and the transaction takes place on an American server situated in the US.
What all of this means to Indian companies is this. Take any website with a dot com extension. http://www.indiainfo.com for instance. Current domain name nomenclature makes it obvious that the content on the site, intended for an Indian audience living in India, resides on a server in the USA.
The content managers have the option of putting it up on a domestic server. The domain name would then read http://www.indiainfo.co.in.
Theoretically, it makes sense to host the content on a local server because traffic would take a direct path. This should translate into a more efficient path. In reality though, it doesn't work that way. In a filing to the Asia Pacific Economic Council (APEC) on these costs, AT&T proved that it costs more to lease intra-Asian circuits than it does to purchase circuits from US operators. Therefore, there is a very strong financial incentive for ISPs to take their traffic to the US and interconnect there. No wonder European and Asian ISPs say that up to 75 per cent of traffic first goes to the US before being routed back.
If this has to change, governments must play a role in infrastructure development. India, for instance, has long been a bandwidth impoverished nation. There has been a flurry of recent activity though. VSNL now owns 165 mbps which includes the 45 mbps it recently purchased from FLAG. Then there is another 155 mbps on the cards from SAFE. This should almost double VSNL's capacity to 320 mbps. There are unconfirmed reports that it is signing up for 155 mbps on the Japan-USA line. All this because private ISPs need bandwidth desperately and VSNL is their only option. In the same breath, it ought to be mentioned that VSNL has been making all the right noises about evolving into a regional hub for routing for domestic traffic and that which originates in neighbouring countries. The bigger Indian ISPs like Satyam Infoway and Bharti BT are setting up gateways which will make them independent of VSNL. Others have evinced interest in setting up similar projects. Add to this the fact that for all the global wiring thatis taking place, there are landing points in India. Hence, it looks unlikely that Indians will stay impoverished for long.
But the real issue in all this is about breaking American monopoly and ensuring that money that should logically accrue to Indian players, stays with them. This can happen only when the government steps in, opens up the sector completely and behaves proactively. Else, the danger of India losing out in the long term is very real. It already is.
A case in point is the ERNET which was originally set up with a UNDP grant. It commenced operations much before VSNL got into the bandwidth game, and gave India a head start over China.
Unfortunately though, ERNET is now stuck in bureaucratic wrangles. Its lines are clogged and China has over the last two odd years surged ahead of India both in terms of the bandwidth it has at its disposal and the number of users logging on to the Internet.
Opening up the sector goes beyond lip service and offering incentives to increase regional Internet capacity. Tax breaks or subsidies to backbone providers who deploy high speed local access technologies, should perhaps be considered. The government should consider getting into venture capital "like it is being done in Germany" so that investments in strategic areas is not stifled.
Alliances with neighbouring countries ought to be pursued actively. In Asia there are five backbone networks right now " the A-Bone in Japan, Hong Kong Telecom's Net Plus, Kokusai Denshin Denwa's Internet KDD, Singapore Telecom's Singtel, and Telestra's Big Pond. None of them interconnect with each other.
But there's a frightening flip side to it. In the process of opening up, the sector may end up being over-regulated. Which is why, it becomes important to create a public sector " private entrepreneurship model of the kind which the US pursued in the early nineties. It is also working hard at future proofing its networks by funding research into next generation networks. There is the vBNS (very high speed backbone network service) being developed in association with MCI WorldCom, the Internet2 project which has roped in universities, and Project Abeline maintained in part by Qwest Communications and Cisco Systems.
Obviously, there is a lesson to be learnt. The US has got it right and emulating it would be wise. The other option once again, is to be fossilised.