The idea behind this column is to take a look back at the country’s broadband performance and to focus on a couple of issues of interest; for example a particular ISP’s performance, a particular performance issue, international comparisons or new technologies.
As it’s the first column, I should probably provide a brief explanation of Epitiro’s methodology. The most important thing to remember is that our probes do not sit in the network we are measuring. They sit on the network edge, like a customer, and measure many of the things that make for a good or bad service experience.
We do a wide array of tests in two different ways:
• One of each, every 15 minutes, 24 x 7 x 365, at 11 sites around the country, spread across New Zealand’s five largest cities, and in various other countries around the world, including Australia, the UK and elsewhere
• Daily/twice daily tests from a desktop agent that runs on users’ PCs, in approximately 50 countries around the world.
As Epitiro has expanded its global dataset, it has been able to make some interesting comparisons between New Zealand and other countries’ overall broadband performance.
There is one notable difference between New Zealand and just about every other OECD country, which cannot be explained away because of our lack of a significant cable modem footprint or our embryonic ADSL2.0 presence — two factors that account for a large part of the gap between us and the UK — and that is performance variability.
Take a look at the graphs below. Australia’s pattern since 2003 is fairly typical for many OECD countries; a slight increment in 2006, which is sustained through to the present day, but notably, with little in the way of major fluctuations. Now take a look at the New Zealand graph.
This level of variability — coupled with the low transmission speed — is unique in the OECD. This data comes from our network of desktop agents that sit on users’ PCs. I have used transmission speed here as one example.
If we then turn to the data from our lab sites, where we measure ISP performance under controlled conditions every 15 minutes, we see the same thing, but in New Zealand, also on a daily basis. New Zealand’s performance — to a greater extent than any of the other countries we measure — suffers daily during peak hours.
Conclusion: traffic management and network saturation are occurring. We track management, so can get a pretty good idea of the extent to which this is a factor, and there’s no question it is happening. But it doesn’t account for the level of variability we are seeing.
No country we monitor can match New Zealand for the heights of its peaks or the depths of its troughs, and the frequency with which they occur.
Moving forward, I hope to be able to use this column to integrate the insights from this data not only with our global dataset, but also into some of the other work being conducted by organisations in the same field. For example, it may well be we can provide some quantifiable dimensions around the New Zealand Institute’s “DSL Wall”; a concept that neatly illustrates the limitations of a long-term, copper-based broadband strategy.
Moreover, as the Commerce Commission publishes the quarterly index of broadband performance based upon Epitiro’s data, I hope to be able to talk you through some of the messages in the data via this column. The first report, just published, is a valuable stake in the ground, demonstrating where broadband service levels are at, prior to the effects of regulation becoming manifest. Over the next few years, we hope to see a substantial improvement, as LLU brings ADSL2.0 kit into local exchanges, mobile broadband closes the gap with DSL and Telecom Wholesale offers FTTN-based services.
See you next month.
Cranna is managing director of broadband benchmarking company Epitiro Technologies in Australasia.