news A series of new speed tests on Optus’ fledgling 4G mobile network in the Hunter region of NSW appear to show the network has the potential to be at least as fast as Telstra’s 4G network, reversing earlier results which had appeared to show Optus’ infrastructure was much slower than that of its rival.
Several weeks ago, a real-world comparison test conducted by Lifehacker purported to show that Optus’ 4G mobile network, which has so far only been rolled out in Newcastle and the Hunter region of NSW, does not deliver anywhere near the speeds possible on Telstra’s rival 4G infrastructure, despite claims by Optus that the telco was planning to build “the best-performing network in Australia”.
When the telco announced its Newcastle 4G rollout, Optus was believed to have been seeing download speeds up to 50Mbps on the 4G network, with upload speeds up to 20Mbps. It recently said the network was capable of typical download speeds ranging from 25Mbps to 87Mbps. This — on paper — compared favourably with Telstra’s network, which is theoretically capable of download speeds between 2Mbps and 40Mbps, and upload spees between 1Mbps to 10Mbps. Recent real-world tests using HTC’s 4G One XL handset on Telstra’s network in Sydney have shown speeds of around 35Mbps down and up between 15Mbps and 25Mbps.
However, real-world tests conducted in the Hunter region by Lifehacker of the 4G networks of both Optus and Telstra consequently showed that Telstra’s network was consistently faster — regularly achieving download speeds in excess of 30Mbps in the region, compared with Optus’ network, which topped at at close to 14Mbps. We recommend you click here to check out the full test results.
Over the past several days, however, new evidence has arisen that shows Optus’ 4G infrastructure has the potential to at least match Telstra’s network – or even exceed it in places.
Firstly, in an article published on July 9, technology magazine TechLife provided a forward highlight of the results of its own tests of Optus’ 4G network, noting that while Telstra still had “the 4G crown”, Optus’ 4G infrastructure was “neck and neck” with Telstra in most of the magazine’s testing. “The basics look very good, especially considering Optus is still in “trial” phase,” the magazine noted.
In addition, Whirlpool user ‘breaker’ has published an extensive series of tests of the network, including screenshots from the Speedtest.net site regularly used to test broadband networks in Australia. The tests, ‘breaker’ wrote, regularly showed speeds between 20Mbps to 30Mbps, with some built-up areas showing lower speeds between 10Mbps and 18Mbps. These speeds are comparable with the performance of Telstra’s 4G network in tests done by Delimiter in Sydney.
However, the news also gets better for prospective Optus customers. ‘breaker’ also published some tests showing that Optus’ 4G network speeds can regularly range up to 60Mbps – which is significantly faster than Delimiter has been able to achieve on Telstra’s network. In addition, latency on the network (commonly referred to as ‘ping times’ was sufficient to allow the tester to play online games such as Battlefield 3. “Had a game of BF3 with the lads for over an hour, no issues,” ‘breaker’ wrote.
There are a large number of factors which could affect the differing speeds between the two networks.
Optus’ 4G network in Newcastle is based on equipment from Huawei, with the telco reported by a number of sources to be currently evaluating bids from the Chinese company and a number of other suppliers for its full 4G rollout, while Telstra has long standardised on equipment from Ericsson. It’s not known which supplier has the more powerful network equipment in practice, although both Ericsson and Huawei rely on the same wireless standards for 4G services — and the 4G networks of both Telstra and Optus use the same 1800MHz spectrum band.
Other factors may include the quality of the receiving equipment. Both Telstra and Optus are believed to use mobile broadband dongles from Sierra Wireless, although Telstra also has a number of other 4G handsets available from major vendors such as HTC and Samsung.
Another factor is the quality of the backbone links to mobile towers. All of Australia’s mobile carriers have placed a large emphasis on linking mobile towers to fibre backhaul links. However, Telstra has a large amount more infrastructure available than Optus and Vodafone, as Australia’s former monopoly telco.
Lastly, Telstra’s network is in production status, with the telco having connected several hundred thousand live customers to it nationally, while Optus is still testing its 4G infrastructure in the Newcastle and Hunter region as a precursor to deploying 4G more widely in the second half of this year.
opinion/analysis
Very interesting – we’re currently seeing wide variances between the speed tests coming out of Optus’ new 4G network in Newcastle and the Hunter region. One publication claims the speeds are inferior to those of Telstra’s 4G network, while another claims they’re broadly neck and neck. In addition, Whirlpool poster ‘breaker’, who appears to have done the most widespread testing of the network, hasn’t compared it with Telstra’s network in the same areas, but has shown consistently better results than we’ve seen from Telstra’s 4G network in Sydney.
So much is in flux at the moment with this infrastructure – not least the fact that Optus hasn’t yet connected many 4G customers to its new network infrastructure. It will be fascinating to see how the new network performs as things stabilise. If Optus can consistently maintain the results which ‘breaker’ is seeing, it will be very good news for Australia’s many mobile broadband users indeed. We may finally be seeing a mobile network to rival Telstra’s Next G infrastructure – and likely at a cheaper price.
Optus chief Paul O’Sullivan’s claim that Optus is planning to build “the best-performing network in Australia” may have some merit after all. If so, we’ll be applauding the SingTel subsidiary’s efforts.
How are the results to far appart? Is it that something has been changed in the optus enviroment? I’d like to see LifeHacker have a do over.
The difference in speed tests between an operational network and one that is being trialled is hardly surprising. Comparing the two is a bit like comparing apples and oranges.
We will only know if there is any difference between the two networks when we can compare their operation under completed deployment and under full load. Until we can do this any comparison is meaningless.
Comparison is not meaningless. It provides evidence about theoretical speeds. I will not tolerate any future comments along these lines. Renai
That’s a bit harsh Renai.
How do you know that what is being tested during the trial is theoretical performance?
I’m no expert or Optus fan but there are a number of things I imagine they might want to simulate by adjusting their equipment before they rollout Aus wide, and the upper bounds of performance is only one.
I’m not going to tolerate debate over whether comparisons are legitimate. The topic is a matter of public interest. Cheers.
Apart from what Renai says, whats wront with comparing the speeds under Optus’s trial with the speeds Telstra delivered in the early days of its rollout?
We compare speeds of our network versus the other two all the time. We need to know where to improve, speed / reliability comparisons are the easiest way to do this.
While I agree that testing vs active isnt the ‘best’ comparison, its certainly useful – even if its not predictive of the final product. Huawei is constantly testing the network between Optus and Vodafone to determine where (as site engineers – or in my case network engineering) we need to pay attention to.
To dismiss it entirely would be to the detriment of the whole process.
We test during the actual testing phase just as much as the release phase.
We will only know if there is any difference between the two networks when we can compare their operation under completed deployment and under full load. Until we can do this any comparison is meaningless.
This right here is exactly the problem, one is a production network with the customer traffic, the other is an in testing network.
The first test from the other article may easily have been done when Optus was doing it’s own load testing, hence less available bandwidth for any other services and a slow comparable result, this test may have been run when there was pretty much no one else using the network, hence plenty of bandwidth available giving a faster comparable .
Note I’m here trying to defend Optus or Telstra, I’m saying that the testing methodology is flawed and until both networks are in production any tests done to compare them is pure speculation.
I’m not sure what all the fuss is over – 4G is 4G, the only difference is the amount of spectrum allocated.
‘4G’ is certainly not 4G. The term is a marketing term used by telcos to describe their network speeds. It does not describe the actual standard being used, the backhaul to the towers, or any other influential factor which comes into play when building a 4G network. In fact, some companies have used the term ‘4G’ to describe services which most people argue aren’t actually 4G.
“4g” is about as much a standard as “sandwich” is a standard.
Would like to add to this.
In America, what we know as 3.5G (hsdpa+) is often called “4G” (HTC labels it as “H”, in Canadian/American phones the little label often says “4G” where ours would say “H”)
Nowadays American telcos often label what we call 4G “4G LTE” or just LTE.
I remember reading something that indicates that what we call LTE shouldn’t even be called LTE, since it doesn’t reach some magical number (was it 100/100?) that was ratified as LTE… (though that might have been the 4G moniker, Wikipedia knows more).
Basically, 4G is nothing more than “Current Tech + 1”. It is always worth comparing what each company is offering as their “Highest” tier of technology. This makes them both strive to do better by everyone. In fact, I am sure by highliting the deficiency in Optus’s network, sites like Delimiter etc. have prompted Optus management to push their engineers harder. And or spend more money in an attempt to prove these statistics wrong.
I highly suspect all the attention on the Optus 4G rollout, has if anything improved the quality of the trial sites (resulting in the disparity between current numbers).
@PeterA
That’s not quite true. What we have here, LTE, as compared to LTE-Advanced, has been recognised by the ITU as 4G. Originally it was not, in 2008, when the 4G standard was put forward. But it is now defined as “a substantial level of improvement in performance and capabilities with respect to the initial third generation systems now deployed” http://www.itu.int/net/pressoffice/press_releases/2010/48.aspx
LTE-Advanced and WiMAXX2, which give stationary theoretical, LOS, single user per cell speeds of 1Gbps are “full-fat” 4G if you like. But LTE IS 4G, despite only achieving 1/10th of that (100Mbps theoretical). HSDPA however, is not. Because I don’t think you could call a jump from Real-world speeds on 3G of 3-4Mbps to 8-9Mbps on HSDPA “substantial improvement” in comparison to even LTE’s 15-20Mbps real-world.
But the US companies can get away with it because of the indefined “substantial improvement”. Whereas we’re a little more definitive here, what with the ACCC and all.
So Basically what you’re saying is HSPA technology will eventually become 4G as release have defined Quad-Cell MIMO HSPA which is about 168 Theoretical so you should be able to pull about 50Mbps with load conditions. Then there is Octal Cell and more to come. What they have done is open the door up to call 3G technology 4G.
@SW_Victoria
Not likely. Do you know of any plans by Telcos here to go quad or octal? No, because LTE is more efficient in spectrum for capacity balance.
Telcos can only gain benefits for customers by quad or octal cell HSPA by increasing backhaul to the tower. Why would they do that just using 3G when 4G gives other advantages as well AND doesn’t compete with limited 3G spectrum?
Ok, point taken.
Should add when I said 4G I actually meant LTE. (ps why is there no edit button?)
Sorry Paul, that is way off the mark. You could well say the same about 802.1x wifi, or even ethernet, but again you’d be incorrect. Certification to a standard simply means you technology meets the requirements set by said standard, but there is often tremendous variation in the technical approach to achieve that outcome, and such variation can result in anything from speed and range differences to compatibility issues between vendors, and certainly differences in reliability. You can even have substantial differences in performance from different versions of the same model, so thorough, ongoing testing is crucial to ensure educated decisions are being made about planning, deployment and maintenance.
Oops, should have replied to the comment, not the article…
And as others have pointed out, ‘4G’ is not a technology standard per se, anyway
Well Ethernet isn’t even “broadband” in the pure technical sense; it is “baseband”. Since gigabit ethernet is known as “1000BaseT”. True broadband uses a “broad” (wide) range of frequencies in the “band”.
how many more people are using the telstra 4g v optus 4 when the real world speedtest were done?? we then would have finally a real answer!!!!!!!!!!!
There is under 2 people per base station on Optus Newcastle LTE network. How does this compare with Telstra?
Thats awesome that some of my results have been published!
I have been truely impressed with the Optus 4G Network.
The trial ends soon 5th August.. so iam told, so it wont be long till it is released to the public, which will be interesting with more People/Load on the cells.
I ran another speed test at home here in Cameron Park, 2285 which is a bit of a broadband dead zone, with friends and family unable to even get ADSL1.
http://www.speedtest.net/result/2084782449.png On a thursday night about 6.50pm peak time :)
Pricing and plans will be the next focus.
Breaker…
“Comparison is not meaningless. It provides evidence about theoretical speeds. I will not tolerate any future comments along these lines. Renai”
Wow here we go. Maybe you should title your article ‘Optus trial 4G theoretically just as fast as Telstra’s 4G”
If we are going to test theoreticals you might want to mention Telstra theoretical max achieved over 140mbps down.
I’m just waiting for the I told you so moment to come where optus’ 4G flops exactly like its 3G network.
Optus you should be focusing on your crap 3G network.
Comments are closed.