Where is the global internet headed and how is it going to get there?
There are a few ways to answer that question.
Big picture: the combined effects of new internet-enabled devices, growing broadband penetration in developing markets, higher broadband access rates, and bandwidth-intensive applications continue to fuel strong internet traffic growth.
Outside of some developing countries, the days of triple digit annual growth rates are long gone.
And while end-user traffic requirements will continue to rise, not all of this demand will translate directly into the need for new long-haul capacity. Aggregate international capacity and traffic growth rates are slowing as the global internet matures. Outside of some developing countries, the days of triple digit annual growth rates are long gone. A variety of factors shape how the global internet will develop in coming years.
IP backbone operators must make considerable investments in network capacity to keep up with rapid traffic growth, driving concerns that capacity costs will outstrip traffic revenues. As IP transit prices have eroded, the convention of expressing transactions as unit price per Mbps yields conspicuously low figures, which contributes to a sentiment that the transit price trajectory is unsustainably cheap. But as aggregate volumes increase and prices fall, a parallel effect takes place on network cost. High-capacity ports afford lower unit traffic costs than low-capacity ports.
Technological and economic advancements in transport infrastructure also have lowered capacity costs for IP traffic. For example, 100 Gbps transmission can transport traffic at a lower cost per bit than 10 Gbps. Although 100 GigE ports pose higher cost and traffic topology hurdles at layer 3 than at the transport layer, the industry will eventually overcome these barriers, yielding continued reduction in cost per bit and greater efficiency in network architecture.
It is not a bold prediction that IP transit prices will continue to fall globally, as they always have. The rate of decline will be greatest in emerging markets with high prices that have greater potential to fall due to increases in volume and local traffic exchange that improve economy of scale.
Prices will fall in the established hubs of internet traffic exchange, enabled by escalating volume and declining unit cost.
Prices will also fall in the established hubs of internet traffic exchange, enabled by escalating volume and declining unit cost. Although 10 Gbps price declines are mitigating in major hubs, the shift to 100 Gbps ports will usher in another step in aggregate unit price decline.
The increased reliance on direct connections to content providers and the use of caching will continue to have a localizing effect on traffic patterns and dampen international internet traffic growth.
While the increase in broadband users and access rates will continue to drive traffic growth in access networks, much of this growth may be managed locally.
As demand for content and applications grows in new locations, CDN nodes and caches will be deployed into these areas. While the increase in broadband users and access rates will continue to drive traffic growth in access networks, much of this growth may be managed locally within a network and may not lead to proportional increases in traffic on international links.
The largest content providers’ private networks are having a major impact on the growth of internet capacity requirements, particularly on the highest capacity routes.
As the content providers extend their networks into new locations, backbone operators optimize their topologies in response. As a result, the pace of international internet capacity and traffic growth appears likely to continue a pattern of broad deceleration.