MyBroadband gets an 'F' for fail

The Minister for Communications Malcolm Turnbull provided one of the missing pieces of information in the NBN debate this month with the release of the complete Broadband Availability and Quality Report 2013 and the launch of the Department of Communications MyBroadband website.

The MyBroadband overview states “to assist with the prioritisation of under-served areas, the Government asked the Department of Communications to undertake an analysis of broadband availability and quality in all areas of Australia.”

It’s a lofty goal and one for which the government should be applauded because the lack of information about the true state of broadband in Australia has hampered the NBN debate.

Unfortunately, the MyBroadband website raises more questions than answers.

FAQ fog

The first hint that something is wrong comes when you read answers to the frequently asked questions.

The data used in the analysis is described as “a snapshot of broadband as at December 2013” and includes:

• Network coverage data from several different telecommunications carriers (current as of September, October and December 2013)
• Data from the Telstra Wholesale website
• Over 20,000 real world observations on ADSL usage and the factors that affect ADSL usage
• Locations and numbers of premises for each distribution area

What this means is the data used was provided by “several different telecommunications carriers”, the Telstra Wholesale website and therefore does not provide a comprehensive picture of broadband in Australia.

What is interesting is the information collected by the Australian Bureau of Statistics (ABS) on Internet Activity does not appear to have been used in the analysis.

If the ABS data was used the government would have found that as of June 30, 2013 about 4.8 million of the 12.4 million internet subscribers in Australia were Digital Subscriber Line (DSL) subscribers.

The MyBroadband website included “over 20,000 real world observations on ADSL usage”. Whoopee! That means the government utilised observations on less than 0.5 per cent of DSL subscribers in the data set.

The description of broadband technologies indicates that Dial-up modem technologies “is a basic technology and is not considered to be a broadband technology. As such, dial-up was not assessed in this study”.

Why not? ABS data shows there were 227,000 dial-up internet subscribers on June 30, 2013. Surely identifying where Dial-up customers are located would assist to find under-served areas in need of prioritisation? Or is the government worried that a review would find that Dial-up internet subscribers are mostly located in regional and rural areas – possibly National party electorates?

The International Telecommunications Union (ITU) definition of broadband can be found in Recommendation I.113 of the ITU Standardization Sector which states “transmission capacity that is faster than primary rate Integrated Services Digital Network (ISDN) at 1.5 or 2.0 Megabits per second (Mbits)”.

Other broadband technologies are listed including ADSL, ADSL2 , mobile broadband, Hybrid Fibre-Coaxial (HFC), Fibre-to-the-Premises (FTTP), Fibre-to-the-node (FTTN), NBN Fixed Wireless and Satellite.

Glaring omissions

Interestingly, ISDN also fails to appear as a broadband technology on the MyBroadband website. And we should remember that in Australia, the government, Telstra and other companies were promoting and selling ISDN basic rate access at 256 kbit/s as a “broadband service” when ISDN first became available. Here is a link to some of the information on the Telstra website about ISDN services available today.

The argument that dial-up internet connections are not broadband and will therefore be omitted is not acceptable when the purpose of MyBroadband is to reflect the results of a study into broadband availability and quality. ABS figures show that two per cent of Australians still utilise Dial-Up internet connections and they’re not all located in one place but distributed widely with some on major city fringes.

A colleague of mine, who until recently lived in the Dandenong ranges on the Melbourne fringe, was only able to connect to the internet utilising Telstra ISDN. His problem was that the length of copper from the exchange to his home meant that DSL would not work and only with the use of an ISDN repeater was the ISDN service made available.

So why has ISDN been omitted from the MyBroadband website? For some, ISDN remains the only way to get internet connection speeds that sit between dial-up and DSL.

Broadband profile

To ensure that people don’t get the wrong impression when you enter your address into the MyBroadband website your results are accompanied  with a warning message. The message clearly states that the ratings shown are “indicative only and may not reflect the actual speed of the specific broadband service for the address entered.”

Indicative of what you might ask?

Well the “Broadband profile for your area” includes three sections. The first is the overall fixed broadband availability, the second is the overall fixed broadband quality and finally mobile broadband.

The overall fixed broadband availability “measures access for each premises in your local area to at least one fixed broadband technology.” It appears that if a listed technology can be accessed from the property then it is listed as an ‘A’, otherwise listed as ‘not available’. If HFC goes past my street or stops 50 meters from my home do I get a ‘B’ or a ‘C’? There is no mention of the broadband availability scale on the website.

The overall fixed broadband quality “describes quality in terms of the speed capability of the fixed broadband technology available to premises in your local area.”

The mobile broadband availability and quality provide information on whether mobile coverage exists and what the highest network type available is, for example “4G coverage”.

Nothing to see here

But it was at this point that it became clear that there was a significant problem with the information shown. Let me explain.

The overall fixed broadband quality section is misleading.

Quality usually relates to several factors, not just the “speed capability of the fixed broadband technology”. But if quality was just limited to the “speed capability of the fixed broadband technology” the website is still misleading and technically incorrect.

The website stated that HFC was an ‘A’ reflecting the DOCSIS 3.0 standard implementation by Telstra and Optus. DOCSIS 3.0 provides for channel bonding to utilise 24 channels downstream and 8 channels upstream, yet Telstra and Optus are currently providing connection speeds below what is possible. DOCSIS 3.0 also incorporates quality of service capability to ensure services such as Voice over IP work correctly, yet during peak periods both Telstra and Optus cable have serious contention problems which negate any quality of service implementation.

So should the HFC rating have been ‘A’?

Not surprisingly the ADSL quality rating was a ‘C’ but on what scale? ADSL is not implemented in Australia with quality of service, capacity and traffic class management capability. Traffic is sent utilising “best effort” transmission which provides the lowest quality of service capability.

The quality rating should be provided relative to the best technologies capability and implementation. Currently this is FTTP and NBN Co has commenced offering a FTTP wholesale product with 1 Gbps/400Mbps connections with quality of service, traffic and capacity management from January 2014.

Utilising the FTTP capability and implementation as a baseline, and selecting the Optus HFC network because it is implemented at 4 channels down and 1 channel up (152 Mbps/27Mbps) which appears to be lower than the Telstra HFC network capability of 8 channels down and 4 channels up (400 Mbps/108 Mbps), the rating for HFC should have been 152 out of 1000 for download, 27 out of 400 for upload and 0 out of 3 for quality of service, traffic and capacity management.

The result for ADSL (in my area) would be 12.12 out of 1000 for download, 0.8 out of 400 for upload and 0 out of 3 for quality of service, traffic and capacity management. A quick series of speedtests during the busy evening period found my home ADSL service provided about 8 Mbps down and 0.6 Mbps up.

The Department’s report, which did not provide any mathematical equations, states regarding HFC quality:

While HFC networks support services with peak download speeds approaching 110 Mbps, these networks were weighted lower than FTTP due to the lower upload speeds available and the current contention ratios for shared infrastructure which can limit the potential for these networks to deliver high speed services to a large number of premises simultaneously.

So how on the MyBroadband website can FTTP and HFC both be given an ‘A’?

For a location picked at random, 158 Brunswick Road, Brunswick Victoria the ratings were FTTP ‘C’, HFC ‘A’ and ADSL ‘B’. Hang on to your seats! The overall fixed broadband quality rating was ‘A’. How is this possible?

Australia is funding NBN Co to the tune of $43 billion or more to install a new optical access network. According to the MyBroadband website FTTP at 158 Brunswick Road, Brunswick Victoria receives a ‘C’ quality rating.

FTTP is a new standard that utilises an all optical network and if it is installed and operated according to the standard Australians should expect that FTTP will always get an ‘A’ for quality.

'F' for fail

At the Senate Estimates Committee on the NBN last week, the NBN Co Chairman and CEO Dr Ziggy Switkowski complained that he was spending too much time in front of the committee. Possibly he is there for good reason and at his next appearance the question should be put to Dr Switkowski about why NBN Co is installing FTTP and it is receiving a ‘C’ quality rating.

Without knowing the mathematical equations used in the report it is hard to judge the validity of the report’s findings but if the report is reflected in the MyBroadband website then serious questions need to be asked about the methodology and the findings. The mathematical model must be released for public scrutiny. Samples taken randomly from the data set should be released so the model can be tested.

A layman will be confused by the MyBroadband website into thinking that an ‘A’ means they can access the best quality broadband access network currently available in Australia yet this is not the case.

Or has this been another carefully crafted exercise to give people an impression that HFC means ‘A’ and that is all people need?

The coalition government does not help its cause by serving up gibberish that does not provide a clear and understandable statement of the true state of broadband in Australia. My rating for the MyBroadband website is ‘F’ for fail. Turnbull needs to remove the website until the deficiencies have been fixed.

Mark Gregory is a Senior Lecturer in the School of Electrical and Computer Engineering at RMIT University