The maximum possible speed of an ADSL2+ service will depend on a number of factors. Some of these - such as the length and quality of the copper wire from your place to the local exchange - are outside of your control. However many others can easily be mitigated:
- the quality of the cabling within your premises - a central splitter is the best way to address this issue;
- the ADSL 'line profile' used on your service - this can be adjusted to achieve optimal performance for your specific circumstances; and
- the capabilities of your ADSL modem or router - all routers sold by Internode are fully ADSL2+ Annex M capable.
How fast can it be?
On the graph, the blue and green shaded areas show the theoretical maximum speeds you can obtain on Internode Extreme, Easy Broadband and Easy Naked services (when delivered via an Internode DSLAM). This shows the effect of distance from your local telephone exchange, and the use of either an ADSL1 or ADSL2+ broadband modem.
Figure 1: Graph showing the maximum speeds in relation to distance of ADSL1 and ADSL2+
Factors that may affect performance
While the speeds indicated on this graph show the theoretical maximum speed of the various ADSL protocols on good copper lines - with no bad joints, faulty insulation, or high speed 'interferers' (such as bridge taps or loading coils). Your actual achieved speed may vary substantially from these theoretical results, based on:
- the 'line profile' that is operative on your service.
Internode places new Easy Broadband services on a conservative initial profile that constrains maximum download speeds, in order to achieve the highest possible connection stability. Customers may then change (online) to any other line profile, at any time, and at no cost, to explore the tradeoff between speed and connection stability that applies for their individual situation.
- the length of copper wire from your premises to the exchange - this is always higher than the distance 'as the crow flies'.
- the number and type of other digital services being used by other customers over copper pairs within the same cable sheath.
- the configuration and line quality of the copper wire between the exchange and your premises.
- electrical interference from outside sources (such as electric motors).
- the configuration of the copper wiring within your premises (i.e. use a central splitter for optimal performance).
- the software configuration and application on your computer (in particular how it uses the uplink back to the exchange).
- your ADSL hardware.
- the capacity of, load on, and access data rate of the destination host computer which you are accessing.
The most effective way to improve ADSL2+ performance is to have a licensed cabler install a good quality central splitter (available from Internode) in your premises.
More information on optimising your Internode ADSL performance
So how does ADSL2+ perform in the real world?
You don't live in a laboratory, so it's quite true that the various factors listed above will reduce your performance from the 'ideal' shown in the above graph.
Hence, in December 2006 we took a random sample 7,305 Internode ADSL2+ broadband services, each connected to an Agile DSLAM and using the ADSL2+ protocol (G992.5 Annex A ADSL2+ over POTS). We found the following distribution:
- 13.4% achieve a download synch speed of higher than 20 Mbps
- 27.7% achieve a download synch speed of between 15 Mbps and 20 Mbps
- 22.1% achieve a download synch speed of between 10 Mbps and 15 Mbps
- 23.0% achieve a download synch speed of between 5 Mbps and 10 Mbps
- 13.8% achieve a download synch speed of less than 5 Mbps
So over 63% of these customers were achieving 10 Mbps download synch speed or better.
ADSL2+ Annex M increases the maximum potential upload speed on an ADSL2+ service from 1 Mbps to 2.5 Mbps, enabling high performance Internet Virtual Private Networks from your office.
The above notes on factors affecting your actual performance are equally true for Annex M - particularly distance and premises cabling.
Once again, the 'real world' data is quite impressive - during December 2006 we took a random sample of 401 Internode ADSL2+ services, each connected to an Internode DSLAM and using the ADSL2+ Annex M protocol (G992.5 Annex M ADSL2+ over POTS). And we found the following distribution:
- 10.5% achieve an upload synch speed of higher than 2.4 Mbps
- 24.7% achieve an upload synch speed of between 2.1 Mbps and 2.4 Mbps
- 18.5% achieve an upload synch speed of between 1.8 Mbps and 2.1 Mbps
- 40.4% achieve an upload synch speed of between 1.0 Mbps and 1.8 Mbps
- 6.0% achieve an upload synch speed of less than 1.0 Mbps
So over 53% of these customers were achieving 1.8 Mbps upload synch speed or better.