Are IXPs the turning point for content development in the region?

“IXPs” or “Internet Exchange Points” is one of the latest buzzword in the Caribbean. Pundits promise that it will revolutionise Internet development in the region, but what is it? And is all this excitement truly warranted?

Consider this: Two separate Internet networks (or ISPs) are located in relatively close proximity to each other. Normally, all of their data is transmitted via the Internet, regardless of destination. They each purchase capacity on submarine cables or via satellite to get the traffic off-country, and they pay international transit and third party providers to have the data delivered to the specified destinations.

Consider also that for both of those networks, a fair amount of their traffic is destined for the other. It is still necessary for traffic intended for a local address to be sent internationally to the Internet, only to be returned back to the country, and possibly to someone in the neighbourhood or even across the street from the sender? The short answer is no. Depending on the amount of traffic, a more economical alternative might be for the ISPs to connect their networks directly, and route locally destined data through that link.

An Internet Exchange Point (IXP) is a more elaborate version of this concept. It is infrastructure through which ISPs exchange Internet traffic via mutual peering agreements, to allow traffic to be exchanged without charge.  The IXP allows a variety of Internet networks and providers, such as ISPs, telecoms, broadband, mobile and cable operators and even large corporates, to connect directly with each other and exchange data.

In establishing an IXP, all Internet traffic does not need long distance links. Traffic exchanged via the IXP is usually not billed, and in reducing the amount of traffic that must be handled by transit and third party providers in turn decreases the per bit delivery costs payable by the local operators and providers.

Another advantage of having an IXP, is that delays, e.g. in delivering emails and loading up local web pages, which is usually referred to as latency, can be significantly diminished for data delivered via the exchange. Hence there can be an improved Internet experience, which seems faster, more efficient and more responsive.

Finally, many of us would have realised that rarely do we experience anywhere close to the upper limit of the connection speed that we pay for. For example, if you have subscribed to an Internet package that offers a download speed of up to 1 Mbps, rarely will you connect at that speed, or even 75% of that. One of the key reasons for that is that your ISP cannot guarantee off-country connection speeds. At peak times of the day, Internet speeds can be noticeably slower than at others. However, for services and providers that are connected to the same IXP, connection speeds will be consistently faster, than for those that must travel through the Internet.

Want to test your connection speeds?Click here

Currently, many countries in the region are actively considering deploying national IXPs – none exist in the English-speaking Caribbean. They and locally based ISPs are aware of the benefits (discussed above) that could be accrued. However, many of those who are advocating widespread IXP deployment in the region, seem to be suggesting that increased local content and Internet competitiveness will automatically result. Unfortunately, this is unlikely to be the case.

Local webhosting limited and not competitive

In most countries there is a fair amount of Internet content that is being created and hosted locally. Key sources would be Government, academic institutions, indigenous banks, utility companies, and large local organisations for which having in-house web servers are an advantage. However, for many other organisations and even for the individual user, locally available web-hosting facilities are limited and not necessarily competitive with that available internationally. Occasionally, such facilities even resell products and services that originate elsewhere. Hence, unless or until local webhosts can offer a broad range of features as well as sufficient protection from loss or failure, all of which is competitively priced, content creators may still opt to have their products hosted overseas, and the anticipated benefits of IXPs will not necessarily be realised.

Questionable RoI

Deploying national IXPs would facilitate low-latency and high bandwidth applications, such as multimedia streaming, video conferencing and interactive services. However, in countries of relatively small populations within the region, hosting those applications locally, and if relying primarily on the local population, might not be profitable or yield an adequate return on investment. This situation might already obtain for organisations or businesses with operations throughout the region. It might not be economical for them to have individual or independent web offerings for each country they serve, when the traffic volumes each generates are considered.

In that regard, there might be a strong case for a regional or sub-regional IXP, e.g. for the ECTEL region (St. Kitts and Nevis, Dominica, St. Lucia, St. Vincent and the Grenadines, and Grenada). None of those countries have a population over 200,000. The smallest, St. Kitts and Nevis, has a population of approximately 50,000, but their combined population is over 500,000. That arrangement would improve both the economies of scale and scope, and would be consistent with the existing premise of maintaining a single telecoms space among those islands.

Limited financial benefits to subscribers and users

One of the key benefits of establishing IXPs is the reduction in payments to transit and third parties for handling Internet traffic. Although network operators and ISPs will immediately enjoy those savings, it remains unclear whether subscribers will experience any reductions in the rates payable for access to Internet service.

Internet providers and policy makers should recognise that the ubiquitous use of mobile phones in the region occurred, thanks in large part to the relatively low rates that were introduced in the early 2000s. Currently, it is the primary mode of telecommunications in the region, and has been the impetus for innovation and content development, such as through the mobile applications development both here and internationally.

Hence, while establishing national IXPs in the Caribbean will yield benefits to Internet operators and providers, and even to users, these systems, by themselves, are unlikely to accelerate local content creation. Considerable and affordable resources are already available and accessible over the Internet, and users pay the access rates prescribed by local providers. Internet Providers, and perhaps policy makers, must still be prepared to offer further benefits to users to increase the local content that is generated and accessible over the Internet.


1 Comment

  • Ideas expounded in this article are wide-reaching, in my view. I like the idea of ‘grouping’ in regional or sub-regional IXP. I believe ECTEL ( though specifically in Telecommunication ) was founded on the same principle.

    Small islands of the pacific and developing countries esp. in sub-Saharan Africa you shouldn’t look farther than here.

Comments are closed.