It’s easy these days to see how fast your mobile network is – just run a test on your mobile device with a service such as Speedtest.net and it will give you measured results.
But – as I expect you would guess – there’s more to it than simply looking at download/upload rates and ping speeds such as you see in the screenshot.
What this shows is the results of occasional network speed tests I’ve run over the past month on my Samsung Galaxy SIII LTE smartphone, which I have courtesy of EE, the UK network that’s notable for rolling out the country’s first commercial 4G service. I’m taking part in an ambassador programme for EE, organized by Andrew Grill.
4G service hasn’t yet reached my neck of the woods – it’s coming this quarter, says EE – so when I’m not in London (one of the 16 UK cities that’s currently got 4G from EE), I’m reliant on the 3G service that every mobile operator offers, when I’m not connected to a wifi network.
So every now and again, I run a test to see how fast things are.
The trouble with testing this way is that you could run another test five minutes after one test, and get entirely different results, better or worse. Much depends on obvious things like where you and your device are: inside a building, out in the open, near a cellular signal mast, etc. Then there’s the technical stuff that engineers worry about. Load balancing on the network, signal strength, how many users are connected to a given cell, transmission issues, the weather, that sort of thing.
And an important note: tests like this mean nothing to voice and text messaging – they’re all about data use by your mobile device.
Still, tools like Speedtest’s are widely available, people will use them and come to conclusions based on the results they see.
One thing I was surprised to see, in most tests I ran when connected to EE’s 4G network, was how upload speed was considerably more than download speed by a massive factor at the times I ran the tests. Great if I were uploading loads of files (as my fellow EE ambassador Paul Clarke often does with large video files); not so good perhaps if I were updating a dozen apps from Google Play or streaming Africa on the BBC iPlayer app for Android (as I did do the other day).
But in reality, my experiences with such things have been great so far. That’s as much a testament to the Galaxy SIII LTE itself and its quad-core processor, Android version, fast graphics, overall memory, optimized stuff for 4G and other hardware and software that make up the device, as it is to EE’s 3G and 4G networks (and the cable broadband internet providers such as Virgin Media I connect to via wifi).
I guess my concluding point is that speed testing of cellular networks in particular isn’t really going to give you insight into, or satisfaction about, your experience at any given moment unless you’re testing something specific – and know how to analyse and interpret the results.
It’s a good-to-know kind of thing, I reckon, useful when you are experiencing something obviously not up to scratch like a noticeably slow data connection and you need some specific metrics for when you call the support number or as a screenshot for a blog post or tweet.
Meanwhile, I’m enjoying my experience with EE so far, even if it’s not yet via a 4G network where I am. Can’t wait for that in my area!
Related posts:
2 responses to “The peculiarities of cellular network speeds”
It’s interesting that ‘ping’ is still used, particularly when not allowing the ping packet size to be adjusted. in most cases the ping size is 64 bits, very small, where as data is usually set to larger sizes, you can ping with larger sizes to detect network contention issues where small packet work fine.
[…] The peculiarities of cellular network speeds It’s easy these days to see how fast your mobile network is – just run a test on your mobile device with a service such as Speedtest.net and it will give you measured results. But ̵… […]