How high-speed wireless compares to cable in boosting our internet speeds


This article is by Thas Ampalavanapillai Nirmalathas, Director – Melbourne Networked Society Institute, Professor of Electrical and Electronic Engineering, Co-Founder/Academic Director – Melbourne Accelerator Program, University of Melbourne. It originally appeared on The Conversation.

opinion/analysis Australia’s national broadband network continues its roll out with more than 900,000 premises now connected, according to NBN Co’s latest weekly progress report.

But Google recently announced the development of high-speed wireless internet connections, which raises the question of which technology is the best for any future broadband network.

Political influence

The technology options for deploying broadband in Australia have reflected political thinking and been heavily influenced by changes in federal government. Broadband technology also keeps improving and the continuing debate about choice of technology is hindering effective investment.

The current option for the NBN is a multi-technology mix (MTM). This approach emerged after a strategic review of the project undertaken by the Abbott-Turnbull Government.

MTM has the NBN network comprising optical fibre to the home, hybrid fibre coaxial (HFC) cable network, hybrid fibre-copper cable and wireless options (fixed wireless and satellite wireless access).

But Google’s wireless plan adds fuel to the ongoing debate about whether cable connections will soon be obsolete?

The original fibre plan

The NBN’s original choice of fibre to the premises (FTTP) remains not only the most future proof against rapid changes in technological evolution but also offers the highest speeds of connectivity and lowest delay (latency).

Low latency is very important for the smooth operation of any interactive application. In fact, telecommunications networks rely on optical fibre and the closer that the fibre can get to end user, the better the performance.

Hence, the debate has centred around how best to connect users to fibre over the last few kilometres.

NBN Co is interested in the HFC cable sourced from Telstra and Optus as it has an installed fibre network base within 10km of the subscribers who are connected by a copper-based coaxial cable. This has greater capacity than the standard twisted copper pair phone lines.

If NBN Co can bring fibre close to a network termination point and then utilise the existing copper based telephone lines, there are new technologies that can push the speeds of connection in the short term to that comparable to fibre.

But their upgradability is low and the last few kilometres of copper as well as the terminal equipment will need to be replaced with optical fibre or other improved technologies in the future.

Wireless still needs cable

Even the broadband wireless options being considered by Google rely on a wireless access point located close to the users. That is almost guaranteed to be connected by the optical fibre.

In order to provide wireless access with broadband, the access point needs to have a directional beam towards the user or have a smaller radius of coverage with fewer users. This would imply a significantly large number of wireless access points interconnected by an optical fibre network to provide a coverage to everyone.

There are plans to provide wireless broadband using satellites (NBN’s Sky Muster and others) and other floating platforms (Google’s Loon Project).

But these cannot offer high bandwidth, as it will be shared by a larger number of users. They would also be significantly limited by the excessive delay caused by getting the signal to space and back again.

There is no one ideal solution, but it depends on the population density. Satellites and other space-based platforms provide universal coverage over the Earth’s surface with reasonable bandwidth in a shared manner. But the delay would be a problem when it comes to highly interactive applications involving audio, video or real-time interactions.

It is ideally suited for hard-to-reach places as well as for providing universal coverage of basic broadband to sparsely located populations.

Alternative wireless technologies, such as those being considered by Google, provide rapid deployment at a lower cost. But they also have a compromise in terms of upgradability.

Similarly, all other hybrid options such as fibre to the node (FTTN), fibre to the building (FTTB), fibre to the distribution point (FTTdP) and HFC provide low-cost activation options while deferring the upgrade costs to a future date.

These are options that network operators have when they are making decisions about investing their capital to expand the network, and shorten the time to bring services to market.

Looming election

Will the looming election cause a further change in the NBN policy?

Rethinks based on political ideology has already caused significant delays and uncertainty regarding the project.

What’s needed is bipartisan commitment to accelerating NBN deployment along with modernising the infrastructure in the core network that will have to support increased access to broadband.

Internet Australia, the body representing internet users, is calling for bipartisan support for any NBN strategy. But it notes that Australia has dropped from 30th to 60th in global rankings for average internet speeds.

So it wants bipartisan support to drop the current MTM strategy in favour to adopting at least fibre to the distribution point, although it says fibre to the premises would be the “ultimate goal”.

By Thas Ampalavanapillai Nirmalathas, Director – Melbourne Networked Society Institute, Professor of Electrical and Electronic Engineering, Co-Founder/Academic Director – Melbourne Accelerator Program, University of Melbourne This article was originally published on The Conversation. Read the original article.


  1. Having fast internet isn’t just about download and upload speed. Latency matters as well. Latency and bandwidth will always be a problem with wireless.

    • Bandwidth is the issue. Latency in wireless can actually be far better than fibre, given that the speed of light through air is around 30% faster than through fibre. I work with a company that has an ultra-low latency point to point radio system that is used in high frequency trading because of the latency advantage.

      It’s a little more tricky in point-to-multipoint (i.e. cellular wireless), and 3G technologies really suffered with the TDMA (time-division multiple access). The frequency-division multiple access in 4G (LTE) is far better now.

      The real problem which makes wireless unsuitable as a replacement for fibre for most people is spectrum limits – with the growth of data we’d need a wireless tower for every couple of houses, and most of them fed with fibre. Even if you did decide to do this, running direct fibre is far more energy efficient.

  2. When I heard of the TPG – Vodafone deal to provide connectivity to mobile towers I did wonder how long before we have micro-cells or some other wireless access method to challenge FTTN. Cleverness with copper is soon running out of road. G.Fast is interesting, but it puts an awful lot of of active kit on the access network and I’m not sure where you go after 1Gbps.

    There is plenty of innovation potential in wireless. Access is going to end up being fibre for fixed line plus variation on wireless including satellite, high altitude balloons, vehicles etc and fixed terrestrial.

  3. Bandwidth may be more or less solved, quality of service is what matters. Wireless is just bad, more power, more exposure to attenuation, multipath and interferance. A glass fibre is a dedicated path with essentially unlimited bandwidth, no interferance and minimal delay.

    Even the NBN fixed wireless is far more limited in bandwidth than it appears at the moment, Google’s project of FTTN+wireless instead of FTTN+copper is going to be unsuitable in dense areas.

    If anyone still thinks that FTTP is some gold plated, super expensive tech, just look at Viet Nam where one can buy unlimited 27mbit/s FTTH for just under $17 per month. That fiber is delivered using the same GPON technology chosen by our NBN.

    • In Da Nang Vietnam I’ve got 1x 10Mb/s connection VNPT $8/m and 1x 20Mb/s Viettel $12/m in Da Nang – both unlimitted -very good speedtests to Australia ie 10/8 and 20/15

      I’ve seen optical fibre everywhere in Vietnam. I’ve got viettel 20/20 in another premises I use in Quy Nhon and on the border with Laos VNPT and Viettel as well as FPT are on every third or fourth power pole. The country must have close to universal Fibre availability by now

  4. I like the comment that it will be slower because of distance. If you understand the speed of light and the distance to the Google loon project you understand that is totally incorrect. in some case it would be faster. but it would be very rare for there to be a noticeable latency increase. Unlike Satellite Loon is very close. and still in the atmosphere so doesn’t take long for a round trip.

  5. Or Japan where you can get 2G/1G unlimited for around $50 per month.

    Or Romania where you can get 1G/1G unlimited for around $40 per month.

    Or Hong Kong where you can get 1G/1G unlimited for around $26 per month.

    Or South Korea where you can get 1G/1G unlimited for around $20 per month.

    I can keep going, but I think you see my point.

  6. Google/smoogle is a complete snow job of obsolete hybrid technologies. One which cannot be upgraded to a faster connection, when heavy congestion occurs.

    Google, is corporation, where all taxes are to be evaded. In the interim the suckerbait customer is always last and is to be milked for the maximum possible profit.

    We are all fully aware that Google Oz, pays next to nothing in local taxes, apart from a tiny amount of unavoidable FBT and GST.

    The Edward Snowden leaks tell us that evil is done daily at Google/NSA corporation.

    After all, a quick check of all Android mobile phone apps on offer. Three out of four contain assorted spyware and other evils, which are not disclosed in the EULA.

    For a smart mobile phone a wise choice would be one that is full open source unlike the closed loop one called Android.

    You pays for what you get and get what you pays for.

Comments are closed.