Woodcock has been involved with the Packet Clearing House, a non-profit institute providing research and operational support for the Internet with routing policy, multi-service transport and economic modelling. Bill is currently in Kathmandu, helping establish the first Internet Exchange Point for Nepal. Gaurab Raj Upadhaya spoke to him about the technical aspects of information technology development in Nepal.
Gaurab Upadhaya: In layman's terms, what is an Internet Exchange?
Internet Exchanges are the "factories" that create Internet bandwidth. ISPs (Internet Service Providers) deliver that bandwidth to consumers. IXes are where ISPs communicate with each other handling customers' traffic with one another. Customers are interested in reaching other ISPs customers so the value of the Internet is created at Internet Exchanges. The job of the ISP is to transmit this bandwidth to the customers. Hauling traffic to and from an IX outside the country is very slow and expensive. Building a local IX like the Nepal Internet Exchange (NPIX) within the country solves this problem by reducing the cost and distance between consumers. This is the market driver for the deployment of broadband Internet access.
And what is the Packet Clearing House?
PCH is a US-based not-for-profit research institute, originally formed in 1993. At that time, 70 percent of the Internet was located in California, but the only IX was located in DC. So 70 percent of Internet traffic originated in California and 70 percent was destined for California. Seventy percent of 70 percent is 49 percent of all Internet traffic, and this 49 percent was travelling an extra 9,000 km round-trip, completely unnecessarily.
Today, about half the PCH's efforts focus on developing nations. In addition to building IXes, PCH also helps to find technological, policy and financial solutions for IXes that have encountered difficulties. PCH also engages in research initiatives like archiving topological changes to the Internet, building economic models and working to improve the scalability and robustness of ISP communications. We are supported by the same community that we work in especially by equipment manufacturers and ISPs.
Why the inertia to start IXes in developing countries?
Really, there is equal need for IXes throughout the world, regardless of the degree of development of any country. However, we've seen a general trend toward stagnation of Internet development in countries that don't have a competitive market environment for telecommunications services. In many developing countries, an incumbent telephone company is in a position which does not encourage Internet market growth, and an IX is a way of reinvigorating that growth. Market reforms which encourage competition can also help bring new ISPs into the mix, bringing consumer costs down and variety of services up.
How has the world progressed on the Nepal Internet Exchange?
NPIX participants have cooperated with remarkable efficiency. They've gone from being competitors to collaborators in the interest of Internet deployment in an impressively short amount of time. Nepal's ISPs have a remarkably clear vision of both the economic reality they face and the best path toward improving their situation. Today, (28 August), we turned on the IX switch for the first time, and there will be three ISPs using the IX to exchange production traffic by this Friday. Within a month, we believe all ISPs in Nepal will be participating in the IX.
Are IXes only meant for Internet service providers?
This was a common misconception for quite a few years, due mostly to ISPs being afraid that customers who come to an IX will want service for free or change ISPs frequently, but what we have discovered in the last few years is that IXes are an excellent place to buy and sell services. Rather then let customers change ISPs more frequently, it allows them to get additional services more easily and also lets ISPs provide services to customers with fewer local provisioning problems.
How can the Nepal government and regulators help?
There are actually a number of critical market reforms needed in Nepal before the Internet will really become available to the majority of the public. Foremost among these are the removal of the constraints which artificially preclude market competition to install fibre infrastructure. One important thing to remember is that the current infrastructure were installed at public expense, it should by all rights be equally available to all members of public.
In fact, however, it is currently being used to exert a monopolistic stranglehold on the local loop delivery of telecommunications services. Three remedies exist. First, the government could ensure that all parties have access to unbundled local loop, while permitting the actual physical facilities to remain under the NTC management. Second, the government could at public expense initiate development of new telecommunications infrastructure and the government could operate in a free and impartial manner making it available to all service providers. Third, the government can simply allow market forces to solve the problems, by deregulating and permitting all competitors to install new infrastructure wherever it is needed. In reality, a combination of the first and third options is probably the most efficient solution.
Globally, where is the Internet headed?
I see a general trend toward mobility and prevalence of small and portable Internet devices like laptops, pagers, cell phones and personal digital assistants. The fact of the matter is that the consumers simply don't wish to be tied to a wired phone, or worry about where to plug-in. This trend towards mobility will be combined with the convergence of voice and data services as we have already seen in hand-held devices that combine e-mail, web browsing and telephony. ISPs will begin to look more and more like cellular telephone providers as their deployment expands to include both wireline service and micro-cellular wireless data service. A prevalence of local Internet Exchanges would guarantee that local communications will be both inexpensive and very high speed.