Home
 » ISP News » 
Sponsored

ISP BT Trials New Way of Checking User Broadband Line Speeds UPDATE

Monday, January 7th, 2019 (10:59 am) - Score 19,336
bt uk isp logo

Consumer UK ISP BT has begun to trial a new way of checking customer broadband speeds, which sounds like it will require special code (firmware) to be added on to the end-users router (Smart Hub etc.) in order to test and monitor their line. This could help them to comply with Ofcom’s forthcoming changes.

Last year the telecoms regulator, Ofcom, announced a new Voluntary Code of Practice for Broadband Speeds (CoP), which among other things would provide UK consumers with more information about their estimated line performance and make it easier to exit your contract if related problems cannot be resolved within 30 days.

Crucially this CoP, which is due to be introduced from 1st March 2019, added a requirement for ISPs to deliver normally available speed estimates based on peak time speeds. Effectively this meant that member ISPs might have to test the actual speeds of a statistically meaningful panel of customers on each broadband package during peak time, which at the time was neither cheap nor easy to do.

Ordinary web-based speedtests are simply not up to the task because they are too unreliable and easily influenced by issues such as local network congestion, limitations of end-user devices and slow WiFi etc. The alternative of using lots of custom routers from SamKnows (Ofcom and Virgin Media use this for their speed reports) has also been rejected by other ISPs as being too expensive or tedious to implement.

Alternatively ISPs have also been exploring different methods, such as the possibility of extracting more performance data from wholesale suppliers (there are some big limitations to what can be done here) or building special connection monitoring code into the end-user’s router (e.g. ASSIA’s TruSpeed). The latter would be difficult to develop if starting from scratch.

NOTE: Putting line monitoring code on to an end-user’s router would only work if the kit is supplied and supported by your ISP, which can change the firmware.

The idea of using custom router code has gained some favour among a few ISPs and BT (possibly including siblings EE and Plusnet) now look like they may be pursuing this method. The ISP has just begun inviting selected customers to help “trial a new way of checking customer line speeds,” which they say will run “tests on your … line to make sure our systems can pick up the new speed checks. This may slow down your line temporarily but it won’t last longer than 24 hours.”

The explanation loosely appears to point toward a system that could rely on some feedback from the end-user’s router, although at this time BT has not explained the specifics or whether they are making use of ASSIA’s new system (we have asked and will report back if they reply). Likewise it’s unclear whether such checks will only be performed once or at regular intervals.

BT will also have to be careful to ensure that any checks do not have a noticeable impact on service speed and latency when the line is most active, which could easily disrupt end-user latency (we suspect they’d try to conduct such tests when the line is idle).

Apparently end-users can also expect to receive an email and text message after the test, which will “indicate the speed changes.” The communication should help to comply with Ofcom’s forthcoming regulation changes. We’ll update again when we learn more.

UPDATE 8th Jan 2019

Just for some context we’ve pasted some of the text from Ofcom’s related High-Level Testing Principles document, which sets out the requirements for how ISPs are supposed to conduct this sort of testing.

Test principles

Each panellist must have a unit capable of running the appropriate test software that can measure download and upload speeds received at the customer premises equipment (CPE). The ISP may run additional quality tests if they wish to do so, but for the purposes of the code measuring download and upload speeds will be sufficient.

The software must perform daily tests for each panellist during peak time (8-10pm for residential services, 12-2pm for business) and the quiet hour (the time at which the ISP expects the network traffic to be least contended). ISPs must ensure that tests are spread out across each of these periods.

To determine the maximum speed achieved on the panellist’s line, ISPs may test throughout the day rather than solely in the expected quiet hour.

The data used to calculate congestion must be updated at least quarterly, drawing on the previous three months’ speed measurements, although ISPs may update more frequently if they wish to do so.

The download and upload speed tests should not be run when user traffic is detected by the unit, as this could result in a negative impact on the performance experienced by the user, and may also compromise the test results.

To avoid detriment to panellists, the data used for the tests must not be included in the panellists’ data allowance (if any).

To assess the capacity of the user’s broadband connection, up to three concurrent transmission control protocol (TCP) connections must be used.

The download and upload tests will consist of downloading and uploading files over a duration of 5 seconds. The application layer protocol must be http 1.1 … The files to be used for the testing will be stored on an Ofcom webserver, from where the ISPs can retrieve them for use.

Meanwhile ISPs with smaller customer bases (generally fewer than 20,000 per product) or narrow geographic coverage are told to discuss suitable alternative approaches with Ofcom, which may include sampling by technology, rather than by product. Very few small ISPs are likely to sign-up to the 2018 code as it would be difficult (technically, practically and economically) for most of them to meet the new requirements.

Add to Diigo
Tags: , ,
Mark Jackson
By Mark Jackson
Mark is a professional technology writer, IT consultant and computer engineer from Dorset (England), he also founded ISPreview in 1999 and enjoys analysing the latest telecoms and broadband developments. Find me on Twitter, , Facebook and Linkedin.
Leave a Comment
41 Responses
  1. Avatar martin

    “trial a new way of checking customer line speeds,” which they say will run “tests on your … line to make sure our systems can pick up the new speed checks. This may slow down your line temporarily but it won’t last longer than 24 hours.”

    No way! Just say NO to BT and Ofcom. Pointless and a waste of time.

  2. Avatar Joe

    Checks would at least solve many problems that go undetected atm and allow OR etc to seek a fix.

    Can’t see a monthly check for a few minutes is going to make the world end!

    • Avatar Tom

      These would be throughput tests. Not testing sync speeds which can already be monitored. Throughput issues won’t be something an OR engineer can fix majority of the time.

    • Avatar ComicBookAssassin

      Nope checks just mean the ISP says that’s the fastest speed and openreach say yes that’s the fastest your line can run it’s as clean as a bell even though it may have a brake down the line. I remember having to fight tooth and nail for the 7th engineer to find the issue then my speed went from 36meg to 70plus meg. I was running on 40meg for about 3to4years then dropped below the minimum and took them to task both my ISP and Openreach as there engineer/phone man just wanted to fault my equipment even though I proved there and then that it was the same on the ISP two routers lol according to the engineer all three where faulty lol, anyway the 7th engineer found the fault near the cab. So doesn’t matter what Speedtests they do if there is no one to accept their is a fault and systematically go through and fix the fault! This means noyhing

  3. Avatar AndyC

    I got this email this morning

    What wasn’t mentioned in the story was the trial is only open at the moment to people with a second line with BT so im guessing thats how they will make sure if it go wrong it doesn’t affect you to bad.

    “What’s the trial is about?

    We want people who have a second line with BT Broadband. It doesn’t matter what kind of broadband you have as long as it’s with BT on a second line.
    We’ll run tests on your second line to make sure our systems can pick up the new speed checks. This may slow down your line temporarily but it won’t last longer than 24 hours.”

    • Avatar Joe

      Given the nature of speedtests i’m struggling to see why a second line matters – they can’t be testing such *!*! code that they fear they will wreck something!

    • BT will probably want to know whether this method actually delivers what they’re expecting in terms of results and having two lines coming into the same property helps set a baseline.

      Mind you in my experience the wiring and extensions on a 2nd line can often be different enough that it may impact speed (I had this at a house I lived at back in the ADSLMax era). I’m also unsure how many individual residential customers would have such a setup, it’s more of a home worker or small office setup.

    • Avatar Joe

      Oh you mean measuring both lines fair enough.

      Yeah I used to have the same with 2 lines on adsl and different speeds which was odd.

  4. Avatar Jim Weir

    There are a number of off the shelf methods of carrying out a CPE side throughput test while dropping all user traffic for the duration of the test (say 10seconds) and this function can be automated quite easily. MLab already offer this with a simple linux integration.

    Letting the end user know is the harder part – because this needs to be done during peak hours, it would be a very fast way of annoying a customer.

    In reality apart from VoIP and such Realtime applications, nearly every other application can buffer over a 10sec reduction without the end user noticing.

    It would also be fairly trivial to saturate the pipe with test data between the end user data and the the max throughout for the line.

    The other question is what are you testing to at the far end – to the ISP core or out to a wider internet destination or just to the nearest CDN?

    • Avatar Joe

      Surely if they are supposed to be giving a ‘real’ consumer speed test it has to be to the net or it won’t be the valid comparison that Ofcom is seeking.

    • Avatar James

      Jim, the methodology is stipulated by Ofcom and does specifically discuss avoidance of end-user traffic as the goal from Ofcom is not to measure local congestion on the end user’s router but to measure network congestion upstream from individual connections (as well as actual performance of the line in “clean and quiet” conditions).

      https://www.ofcom.org.uk/__data/assets/pdf_file/0025/111697/annex-5-testing-principles.pdf

    • Avatar Jim Weir

      @James – thanks for the link – so testing to an independent IX

      The idea of a daily test during peaktime on the sampled lines is problematic – finding an EU idle Ramp Up + 5sec window to undertake this during a daily peaktime period is going to be a problem – Interesting that the testing is using customers with 2nd lines, maybe that needs to be the method across the board, fairly simple to provision sample additional lines for volunteer panelists and provide them with a testing CPE.

      Interesting also that for smaller FTTP ISP’s with geographic coverage they have to test every line not just a sample. That prevents the option of a separate fibre with a test CPE.

    • Avatar CarlT

      That’s not speed testing, it’s reporting latency, packet loss and usage. This actually creates traffic to test the capability of the service.

  5. Avatar JW

    But they already know what speed people are connected at, as the modem says what the line sync speed is?

    Simple solution would to just send a pretend firmware package to the router and see how long it takes?

  6. Avatar Meadmodj

    Surely the focus should be on Quality of Experience particularly on video streams.

    BT already displays the estimated line speed when ordering from historic data. BT (OR) already knows the sync speed of your line and the ISP could easily determine if this is meeting the product that has been sold. Ultrafast or FTTH line speeds should be so high as not to impact normal consumer Internet usage.

    So if we are going to test ISPs surely it should be the experience not just speed. In 2017 I had both BT (40 Mbps) and VM (200Mbps) yet both suffered on a Saturday night on various streams (Now, Netflix, iPlayer etc). Both providers should have been able to provide a single stream to my TV well within their defined products. As always such failures end up with finger pointing and a frantic diagnosis. Freezing streams also cause significant overheads within the network as they try to recover, rewind etc and in some cases we have to resort to rebooting devices to reset. Often other devices were working fine including a PC connecting directly to the same content provider.

    The ISPs can limit their back haul capacity, use CDN in many different ways and may (or may not), have direct network links to the major content providers. Surely what the consumer wants is that they get the speed they contracted for and that their specific internet use will not stall a number of (times film, music etc). Yes Netflix, BBC and others need to monitor their own capacity, which I am sure they do, but what the consumer needs is a set of measures (for streaming, browsing, file transfer) that indicate who is the better ISP at delivering content which is the primary consumer use.

    Surely that can be achieved by network monitoring?

    • Avatar Joe

      “BT already displays the estimated line speed when ordering from historic data. BT (OR) already knows the sync speed of your line and the ISP could easily determine”

      If DSl checker is any guide they can’t get that right for many lines.

    • Avatar Meadmodj

      Current. My experience is that the VDSL figures are reasonably reliable within 10%. ADSL will always be best endeavour but ISPs/OR should be proactive if a FTTC is underperforming. We must stop this “upto” nonsense when Superfast should be greater than 24Mbps with the line should be performing nearer the contracted speed.

      Going forward. My point here is that we are supposedly being sold higher speeds whether Superfast, VIVID, G.Fast or FTTH. Having a measure (particularly an overall average) that says their 60 Mbps broadband is producing 50Mbps or their 900Mbps broadband is producing 800Mbps during peak times is not very useful if the ISP cannot provide a single resilient stream at 5 or 10Mbps on a Saturday night. This will become more acute as we move to 4K and other future services.

      We need good specification kit, QoS settings. sufficient back haul capacity and Internet access in the correct ratios. My view is that Ofcom should focus on Quality being be measured. Greater speed helps to cover up various issues but it doesn’t solve underlying issues.

      If averages are to be used the measures should also be separate for Wet String, Superfast, Ultrafast, Giga so that high capacity lines do not distort the average. (What happened to the Standard Deviation along with the Mean). In future Provider A may be best for Superfast and Provider D for Ultrafast. ISP ranking may also change over time. i.e FTTP Supplier E may be good initially but deteriorate over time as take-up of their network increases. If selecting a product on speed/price the quality of the product should also be visible.

  7. Avatar spurple

    As long as this doesn’t become an excuse to stop me from using a custom router, I’m fine with this :).

    Mind you, not currently a BT customer, but I do use my own home-brew router (with an approved Modem).

  8. Avatar Fabrizio

    I believe Medux.com ( a concern based in Spain ) send boxes out to UK subscribers promising amazon vouchers so they can do speed tests both by ethernet and wifi (the box does both) on behalf of some companies namely Vodafone. I wonder now if that’s why they set themselves up offering this service to UK ISP?

  9. Avatar Chris

    My new talktalk router will display current throughput speeds, it’s shown 29Mbps for the last month I’ve had it, yet I was promised a minimum of 32.5 as pay for the speed boost.

  10. Avatar walkerx

    I’ve got a samknows box installed which provides me with regular information regarding my line, and I get monthly report on it.

    • Avatar Walkerx

      Forgot to add

      It would be good if there was a system that showed estimated speeds across all providers as currently because not with BT they can’t give my speeds or my previous speeds from when with them.

    • Avatar spurple

      Yes, great for you. Getting a sam-knows box is a lottery, so I don’t think it counts as a meaningful way to get more widespread stats.

  11. Avatar Nick sutton

    At the end of the day it’s how long you have to wait for the buffering to complete on a download, I got so fed up waiting that I have now cancelled the land line, and gone 4g mobile,
    So far this is proving very fast and reliable, it also works out at a 1/3rd of what BT were charging me, I can live with 65gb data per month, and a mobile for communication,
    Also having a powerful sim router in the loft I get WiFi reception everywhere,

  12. Avatar ComicBookAssassin

    Nope checks just mean the ISP says that’s the fastest speed and openreach say yes that’s the fastest your line can run it’s as clean as a bell even though it may have a brake down the line. I remember having to fight tooth and nail for the 7th engineer to find the issue then my speed went from 36meg to 70plus meg. I was running on 40meg for about 3to4years then dropped below the minimum and took them to task both my ISP and Openreach as there engineer/phone man just wanted to fault my equipment even though I proved there and then that it was the same on the ISP two routers lol according to the engineer all three where faulty lol, anyway the 7th engineer found the fault near the cab. So doesn’t matter what Speedtests they do if there is no one to accept their is a fault and systematically go through and fix the fault! This means noyhing

  13. Avatar D Knight

    I’d rather get proportional charges based on my bandwidth. Why should I be playing £45 a month for up a 12Mbit ‘broadband’ when 50meters down the road, people can get 80Mbit for £32…!

    • Avatar Joe

      Why should the person 50m down the road subsidise your broadband

    • Avatar SuperFast Dream

      @Joe, it’s hardly D Knights fault that BT haven’t implemented the technology to push that 80Mbps 50 meters further down the line is it.

    • Avatar Joe

      SFD: It may not be 50m down the line. Its quite common for houses even close by to be on different cabinets or substantially different line routes.

      Bt have doubtless made a commercial decision on what to run where. The base cost of maintaining a line and most of the cost of actually providing bb are the same regardless of speed.

      Not sure why he’s paying 45£ -v- 35 for Fttc though. Bt have some special discount (signup) rates for fttc around in the low £30s

  14. Avatar wireless pacman

    I love the Ofcom comment about (a) doing the test during peak hours and (b) not doing the test if the customer is using the service. Seems Ofcom has yet to hear about the popularity of streaming movies!

    • Avatar Joe

      Most people have apps on their pc that run when the system is inactive. Its not technically too hard for it to keep trying on the router for a gap…

    • Avatar wirelesspacman

      Well aware of that Joe. My point was that lots (most even?) of customers these days stream movies etc so the chance of the router spotting a quiet period will not be so great during peak hours.

    • Avatar Joe

      Depends on 3 things; the duration of a test (looks pretty brief on the info) and the window for the test both by peak/off peak and how many days it can occur over.

    • Avatar Jim Weir

      Streaming is pretty bursty traffic – once the buffer fills it isn’t a continuous traffic stream.

      The issue is Real-time applications such as CV or VoIP

    • Avatar wirelesspacman

      Hi Joe, I’d missed the bit about the test only needing to run for five seconds. I can see why (for the reason Jim says as much as anything), though I do wonder if 5 seconds would be enough to give a realistic view.

    • Avatar Joe

      Tradeoffs – You want a detailed test you interfere with the user more. The short test is probably enough for a crude – what speed are you getting test. It will miss some problems a better test would catch

    • Avatar Jim Weir

      As it is regular repeat testing, it will flag capacity issues upstream which really is what Ofcom want to isolate & identify (rather than just Openreach sync as a metric)

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Comments RSS Feed

Javascript must be enabled to post (most browsers do this automatically)

Privacy Notice: Please note that news comments are anonymous, which means that we do NOT require you to enter any real personal details to post a message. By clicking to submit a post you agree to storing your comment content, display name, IP, email and / or website details in our database, for as long as the post remains live.

Only the submitted name and comment will be displayed in public, while the rest will be kept private (we will never share this outside of ISPreview, regardless of whether the data is real or fake). This comment system uses submitted IP, email and website address data to spot abuse and spammers. All data is transferred via an encrypted (https secure) session.

NOTE 1: Sometimes your comment might not appear immediately due to site cache (this is cleared every few hours) or it may be caught by automated moderation / anti-spam.

NOTE 2: Comments that break our rules, spam, troll or post via known fake IP/proxy servers may be blocked or removed.
Cheapest Superfast ISPs
  • Hyperoptic £18.00 (*22.00)
    Avg. Speed 30Mbps, Unlimited
    Gift: Code: SPRING19
  • Vodafone £21.00 (*23.00)
    Avg. Speed 35Mbps, Unlimited
    Gift: None
  • TalkTalk £22.50
    Avg. Speed 36Mbps, Unlimited
    Gift: None
  • Direct Save Telecom £22.95 (*29.95)
    Avg. Speed 35Mbps, Unlimited
    Gift: None
  • Origin Broadband £23.00
    Avg. Speed 35Mbps, Unlimited
    Gift: None
Prices inc. Line Rental | View All
The Top 20 Category Tags
  1. BT (2390)
  2. FTTP (1969)
  3. FTTC (1587)
  4. Building Digital UK (1540)
  5. Politics (1323)
  6. Openreach (1321)
  7. Business (1169)
  8. Statistics (1032)
  9. Mobile Broadband (954)
  10. FTTH (950)
  11. Fibre Optic (933)
  12. Ofcom Regulation (866)
  13. Wireless Internet (855)
  14. 4G (837)
  15. Virgin Media (801)
  16. Sky Broadband (573)
  17. TalkTalk (554)
  18. EE (550)
  19. Vodafone (462)
  20. Security (392)
New Forum Topics
Promotion
Helpful ISP Guides and Tips
»
»
»
»
»
»
»
»
»
»
»
»
»
»
»
»
»
»
»
»
»
»
»
»
»
»
»
»
Sponsored

Copyright © 1999 to Present - ISPreview.co.uk - All Rights Reserved - Terms , Privacy and Cookie Policy , Links , Website Rules , Contact