Telstra hits 450Mbps speeds in 4G trial



news Australia’s largest telco Telstra this week said it had achieved live network speeds of 450Mbps on its Next G mobile broadband network, using the LTE Advanced Carrier Aggregation standard across a combination of the 1800MHz and 2600Mhz spectrum bands.

These speeds, achieved in testing, are several times faster than the theoretical peak network speeds that any mobile carrier offers on 4G today, with the fastest 4G speeds commercially available typically those from Vodafone in some metropolitan areas. The carrier’s 4G network can theoretically range up close to 100Mbps, although average speeds are typically more around the 30Mbps mark, according to recent testing. Optus and Telstra have substantially greater coverage than Vodafone but offer speeds which typically average out slower.

To achieve the speeds, Telstra worked with network partner Ericsson to install equipment for two new 4G Frequency Division Duplexing (FDD) channels of 20MHz bandwidth each on the 2600MHz spectrum band (40MHz FDD), aggregated with 20MHz of 4G on the 1800MHz band. The result was three simultaneous side-by-side paths for the data to travel through to the operational core network. Telstra also used an advanced prototype Cat 9 engineering device that could combine the three channels to achieve these speeds.

Mike Wright, Group Managing Director Telstra Networks said the test was designed to take the network capability for a spin and learn what happens when you lift performance to these types of speeds. “This test allows us to see how the technology works ahead of when we make a future investment in it,” Wright said.

“Conducting this type of test is a significant step in the network engineering and development process. It is essential for us to see how this type of technology works in the live network and understand what needs to be done to continue to absorb the exploding demand in mobile broadband and offer an exceptional customer experience.”

The typical speeds achievable at a commercial level will be lower in practice and while individual users may not consume all of this bandwidth, the ability to effectively triple the typical user speeds possible on Telstra’s 4G service today, means the telco can carry a huge amount of future traffic demand shared across many users.

The test also confirmed Telstra’s ability to successfully bring together three blocks of spectrum to increase the capacity and speeds delivered over the mobile network. Deployment to customers remains a few years away and it is expected the initial implementation may look slightly different to today’s test, with plans to combine one block each of 1800MHz, APT700MHz and 2600MHz spectrum.

”Telstra has the largest holding of APT700MHz and 2600MHz spectrum in Australia and we expect we will eventually be able to offer this service across much if not most of our mobile network footprint, allowing more of our customers to benefit from improved capacity and faster speeds,” said Wright

“More importantly, we are future proofing our network and planning to manage effectively the ever increasing demand for mobile data so that we continue to give our customers access to Australia’s best mobile network”.

Point 1: Sweeeeeeeeeeeet.

Point 2: The caveat: Don’t expect to be getting these theoretical peak speeds consistently in any real-world environment, don’t expect these speeds to be available any time soon, and no, this trial does not mean that fixed-line telco networks of the type represented by the National Broadband Network are irrelevant. It is clear that Australia needs both fixed and wireless networks, as, while they share some uses, they are also clearly complementary.

Point 3: Sweeeeeeeeeeeet.

Image credit: Ariel da Silva Parreira, royalty free


  1. That’s great. Anecdotally, do others think people do a lot of heavy lifting over their mobile broadband connections? My experience is no, but that may be atypical. In any event, I’d like to see if this technology can effectively boost capacity on the mobile networks, rather than speed per connection.

    • Some people rely on 3G/4G as their only access to the internet. I’d expect the popularity of this to increase as 4G deployment covers areas without fibre, which could quickly saturate the spectrum.

  2. A classic Telstra response to VHA threatening their network dominance. Will only make a difference when they have it properly deployed in the network.

  3. Just wait until category 8 LTE arrives.

    You have to ask though what is the justification for the super low caps when they have such high bandwidth available.

    • I believe they don’t have high bandwith avaible, not in the real world.. Contention ratios on mobile towers are very high and that is why they need these restrictive datacaps. So no one actually uses these speeds. I Really don’t see the point in chasing these speed advancements. I would like to see R&D working on ways to imrpove congestion on towers so we can actually start to see practical download limits. I remeber many years ago when I was on 1500/256 ADSL and thought 10GB download quota was a joke. To be able to download at even 100mbps with only a 15GB download limit underminds the point of having speed.

      On a seperate note I would never get a 450mbps connection unless it was prepaid/shaped. Can you imagine the excess on that?

      • “I believe they don’t have high bandwith avaible, not in the real world.. Contention ratios on mobile towers are very high and that is why they need these restrictive datacaps. ”

        Correct. To provide fixed-line level data-caps would literally required *hundreds of thousands* of new mobile towers to be able to cope with the demand.

  4. I fail to understand why they are pushing faster and faster speeds for LTE services when the cost and size of data allowances stay much the same as it always has. 450Mbps is pointless for me, everybody I know, and likely pointless for the rest of mobile internet users around Australia too. All the extra speed does it allow us to hit our monthly allowance in mere seconds, quite literally.

    Personally, I think continuing to spend money on increasing bandwidth without improving data plans is plain stupid.

    • Exactly, I’ve noticed only a marginal improvement in page load speeds over the years using 3G/4G. I wouldn’t dream of trying to stream video on my handset, or shift large files, simply because I would chew through my data allowance in mere minutes. So this is all largely pointless until data limits are increased, and the only way to do this is either allocate more bandwidth (not too likely), or build more towers (again not too likely).

  5. So this test speed was achieved using 3x aggregated 20Mhz 4G channels…

    Are there even any consumer devices out there that can take advantage of the LTE Advanced Carrier Aggregation standard over multiple frequencies?

    And even if they come to market in future, how many channels are available per tower and how much backhaul throughput is available to each tower?

    • Yes Telstra are already selling LTE-A wireless hotspot devices. The LTE-A network is probably faster than the wireless.802.11n within the hotspot.

      This article highlights one of the reasons why I’m a Telstra mobile customer. Massive coverage and bleeding edge tech. Vodafail and their 5 minutes of fame as ‘the fastest’ network. lol. Unfortunately there really is no competition on par with Telstra so I’m at the mercy of their pricing.

  6. So basically these are the peak speeds you can expect at 2:30am if everyone in the area has their phone using wifi, are asleep and you are sitting underneath the tower?

    Gotta love carriers boasting about total bandwidth available from a shared bandwidth medium!

  7. Actually, while I’m no Telstra fan, there is a point to this beyond marketing hype. By increasing potential total performance, stimulating demand for devices than can access these spectrums simultaneously, you do two things – you increase the number of total connections you can have to a single tower (reducing contention, all else remaining equal) and you provide requested data to customers faster, closing their connections sooner, also reducing contention.

    The problem actually comes when users change their usage pattern – if you increase quotas and people start using their phones to deliver more content (such as videos or streaming music) they remain connected and downloading from the network longer, which increases contention. No one thinks quotas shouldn’t increase over time though – even Telstra realise that if they remain uncompetitive with quotas they will eventually lose customers. So technology like this could reduce contention at today’s data demands, but by the time it is being deployed it will, for the most part, simply be keeping up with demand so they aren’t caught out with a ridiculously oversubscribed crippled network (3g anyone?)

    But a press release about early engineering tests like this is pure marketing hype – it won’t be relevant for years. Telstra are trying to show their customers that they are investing in future tech to give the perception of quality and performance, even though today this is essentially vaporware.

  8. Why we’re not talking about Artemis pCell is beyond me. A technology that wipes the floor with anything even theoretical in the next 10 years.

    Nonetheless what is really upsetting is the weak justification for charging such outrageous sums for such paltry bandwidth that can be easily used in a few hours.

    With cells connected with fibre or, ironically, xDSL bonded tails (ethernet over copper), antennas sized reduced (along with lease and legal costs) and with back haul costs being greatly reduced, to less then $20 a mbps and yet even with all of those fundamentalist changes (and with a cost base similar to a normal xDSL retail ISP) the cost of data on a mobile is insanely outrageous.

    Renai how the ACCC allows the industry to have margins (we’re talking costs of 0.000005c per MB) that would make the Mafia blush is beyond me. Don’t we have laws to stop companies from charging such outrageous prices? Why don’t Media ask the hard questions about mobile pricing and why M2M is so expensive. What is the fundamental that causes a $0.30c per minute voice rate, with $0.20c flag.

    Combine this with disgusting levels of over subscription (which a certain amazing provider does) and a move big move from time sensitive applications using a VBR QoS on to general internet usage on a UBR QoS the cost of transmission especially has reduced. Yet we’re still charged as if 100% of our usage on a mobile is ultra time sensitive gold.

    Yes limited layer 1 spectrum reduces the bandwidth available but with the advent of the Artemis pCell any further excuse to us this to justify massive prices is just perpetuation of one of the biggest rorts (and not just in Australia – across the planet).

    • There is more to providing a mobile broadband service than simply how fat the pipe connected to the tower is though. I believe in many cases the main bottleneck these days is the wireless spectrum. This is a fixed limited resource and under 3G many times it is that which is congested. Many times if you can switch to another cell on the same tower you’ll end up with a better (or worse) connection.

      I believe LTE however goes a long way to making this situation better and over time we will see bandwidth caps rise (as we have seen over the last decade anyway). We no longer have to set aside a chunk of spectrum for voice calls. Everything is treated equally spectrum wise with LTE.

      I agree that more speeds have limited use without the bandwidth caps to go with it but there are plenty of uses for a less congested lower latency network even under current bandwidth cap levels (VoIP etc).

      I’m looking forward to when LTE700 is commissioned next year. You’ll be able to use VoIP perfectly well into the bush (much further than the current 3G networks allow normal voice calls even).

    • Hear, hear Pierre! Couldn’t have said it better.

      Theoretical “speed” claims are everywhere already today.

      Try and get Telstra to show you their current network contention stats even today — they won’t — because it would expose them for the lying, misleading and deceptive marketing tactics they use to keep piling more unsuspecting customers onto their already overloaded network. Speed without capacity is meaningless. And with the current growth rates in mobile usage stats, we will run out of meaningful spectrum in the next couple of years.

  9. Man wireless tech is so the future, so glad the Libs got rid of the NBN as we knew it, totally obsolete now!

    • Yeah it’s totally obsolete. It’s not like fibre is already getting 999.6Gbps faster then this ‘future technology’ anyway

  10. So, as I understand this story, the 450 Mbps speed is actually 3 pathways combined into 1. So the real speed cap is 150 MBps, only you get it 3 times.

    Not sure it makes a difference in the real world, as its only the total of 450 MBps that you’d really care about, but it raises questions.

    First, would it be smart enough to detect less used channels and focus on them if needed, second (and related to the first), how does contention effect things, and third, how does it translate to towers and fixed line infrastructure you need to support it?

    First question – if the speed is split evenly across all 3 channels, then you get a three-fold dropoff if just one of the channels is overloaded. Which will happen with every other phone that doesnt use this sort of triple path technology. So it needs to be smart enough to route traffic to the best speed as a priority. Not impossible.

    Second question – will contention just mean everyone in an area gets in each others way, and just ties up contention across 3 bands instead of 1? Same as the first, the risk is hogging bandwidth and getting in each others way.

    Third question – All these mobile signals still need to feed into a fixed line network somewhere, to do the heavy lifting, and connect to the networks. So does using 3 bandwidths mean it takes up 3 fibre lines? I cant see how it couldnt.

    As others have pointed out, the fixed line contention generally means someone is going to hit something like 70 Mbps out of their maximum 100 Mbps in a worst case scenario, but if you’re suddenly talking about (up to) 3 times the lines being used, what does that mean for everyone else?

    None of those questions are insurmountable, but they appear to be in play for everyone if they do become a issue. Either way, if this is done by tripling the lines used, its just one more reason for a FttH build – FttN being built into a tower just wont work.

  11. I’d like to know two things:

    1) How many simultaneous customers will the average tower handle?

    2) How many simultaneous customers do Telstra envisage using such a service network-wide?

Comments are closed.