Virgin Media probed over digital switchover rules

[Java Jive]

Virgin Media probed over digital switchover rules
https://www.bbc.co.uk/news/technology-68255056

"Ofcom is investigating Virgin Media over its compliance with rules to
protect vulnerable customers during the digital switchover.

The telecoms industry is currently changing landline services from
analogue to digital.

But the regulator has concerns about how it could impact some people,
for example those with telecare systems that rely on landlines."

--

Fake news kills!

I may be contacted via the contact address given on my website:
www.macfh.co.uk

[Graham J]

Java Jive wrote:
> Virgin Media probed over digital switchover rules
> https://www.bbc.co.uk/news/technology-68255056
>
> "Ofcom is investigating Virgin Media over its compliance with rules to
> protect vulnerable customers during the digital switchover.
>
> The telecoms industry is currently changing landline services from
> analogue to digital.
>
> But the regulator has concerns about how it could impact some people,
> for example those with telecare systems that rely on landlines."

There's no good reason why a telecare system should not work using VoIP,
but sadly I have found that some don't. So far as I understand they
simply use audio tones for signalling from the base station to head office.

Can anybody explain?


--
Graham J

[Rupert Moss-Eccardt]

Some telecare systems use ABC so have a heartbeat that determines that
the line is up. Indeed, it works even if the line is in use for a voice
call.

[Graham J]

Rupert Moss-Eccardt wrote:

[snip]

>
> Some telecare systems use ABC so have a heartbeat that determines that
> the line is up. Indeed, it works even if the line is in use for a voice
> call.
>

Please suggest a link that will explain ...

Thanks.

--
Graham J

[Mike Humphrey]

On Sat, 10 Feb 2024 15:22:13 +0000, Graham J wrote:
> There's no good reason why a telecare system should not work using VoIP,
> but sadly I have found that some don't. So far as I understand they
> simply use audio tones for signalling from the base station to head
> office.

There's two issues. The first is that a PSTN line is powered from the
exchange - as long as the telecare unit and the exchange are powered on, a
call can be made. And both of these are usually equipped with batteries.
With VoIP, the NTE and router are powered locally, and these are not
usually equipped with batteries. So there's additional kit needed to make
sure it works in a power cut.

The other issue is with VoIP itself. As the name suggests, it's *voice*
over IP - it's not designed to transmit data. The codec may distort tones
to the point they're unrecognisable, and the jitter is likely to be high
enough to mess up high-speed modem signals. Most systems include some
workarounds for this. Virtually every system detects DTMF and converts it
to signalling (i.e. instead of sending the tones for "4" it sends an out
of band signal "dial a DTMF 4", which gets converted back at the remote
end). Some have fax detection which detects a fax and sends that as out of
band data. I don't think any have generic modem detection, though the
public network must do this.

So whether it works depends on what signalling the unit uses. DTMF should
be fine (as long as it doesn't depend on preserving the length of the
tone). Modem signalling might work if the baud rate is low, but won't if
it's higher. And there's hundreds of different types of units, and no
central list of what is compatible.

There are also some units that use out of band signalling (similar to
Redcare for alarms). This means that they don't need a dedicated line, but
they do rely on a copper pair and don't work on VoIP at all.

Mike

[Graham J]

Mike Humphrey wrote:
> On Sat, 10 Feb 2024 15:22:13 +0000, Graham J wrote:
>> There's no good reason why a telecare system should not work using VoIP,
>> but sadly I have found that some don't. So far as I understand they
>> simply use audio tones for signalling from the base station to head
>> office.
>
> There's two issues. The first is that a PSTN line is powered from the
> exchange - as long as the telecare unit and the exchange are powered on, a
> call can be made. And both of these are usually equipped with batteries.
> With VoIP, the NTE and router are powered locally, and these are not
> usually equipped with batteries. So there's additional kit needed to make
> sure it works in a power cut.

I understand that.

> The other issue is with VoIP itself. As the name suggests, it's *voice*
> over IP - it's not designed to transmit data. The codec may distort tones
> to the point they're unrecognisable, and the jitter is likely to be high
> enough to mess up high-speed modem signals. Most systems include some
> workarounds for this. Virtually every system detects DTMF and converts it
> to signalling (i.e. instead of sending the tones for "4" it sends an out
> of band signal "dial a DTMF 4", which gets converted back at the remote
> end). Some have fax detection which detects a fax and sends that as out of
> band data. I don't think any have generic modem detection, though the
> public network must do this.
>
> So whether it works depends on what signalling the unit uses. DTMF should
> be fine (as long as it doesn't depend on preserving the length of the
> tone). Modem signalling might work if the baud rate is low, but won't if
> it's higher. And there's hundreds of different types of units, and no
> central list of what is compatible.

The few telecare systems I've heard seem to use DTMF at a data rate
similar to manual dialling - i.e. about 1 baud. So it's quite
surprising the anecdotal evidence that many don't work with VoIP

> There are also some units that use out of band signalling (similar to
> Redcare for alarms). This means that they don't need a dedicated line, but
> they do rely on a copper pair and don't work on VoIP at all.

Ah! Redcare! Known to stop any form of broadband using the copper pair,
from my recollection.


--
Graham J

[Mike Humphrey]

On Sun, 11 Feb 2024 09:04:54 +0000, Graham J wrote:
> Mike Humphrey wrote:
>> There are also some units that use out of band signalling (similar to
>> Redcare for alarms). This means that they don't need a dedicated line,
>> but they do rely on a copper pair and don't work on VoIP at all.
>
> Ah! Redcare! Known to stop any form of broadband using the copper pair,
> from my recollection.

Yes, as both broadband and Redcare send an out of band signal down the
line using frequencies above those used for voice. It would probably be
possible to arrange for them to use non-overlapping frequencies (at the
cost of the loss of some broadband speed), but the simpler option is to
disallow having both on the same line.

Mike

[Mark Carver]

That's how my mother's one works. Prompted my this thread, I've just
looked up what her company are supplying now, and it's something that
provides mobile phone network connectivity, with ethernet (not WiFi)
back up. Has 48 hour battery back up built in.

[Davey]

And if the mobile system goes down? It did here in East Anglia, during
Storm Bebet.
If the VoIP is down, then the ethernet is probably down too?

--
Davey.

[Mark Carver]

So what would your idea solution be ?

[Tweed]

Once power and comms goes down for a prolonged period the vulnerable need
to be visited and/or moved. A fall alarm doesn’t keep someone warm and fed.

[Graham J]

Davey wrote:

[snip]

>>>
>>> There's no good reason why a telecare system should not work using
>>> VoIP, but sadly I have found that some don't.  So far as I
>>> understand they simply use audio tones for signalling from the base
>>> station to head office.
>>
>> That's how my mother's one works. Prompted my this thread, I've just
>> looked up what her company are supplying now, and it's something that
>> provides mobile phone network connectivity, with ethernet (not WiFi)
>> back up. Has 48 hour battery back up built in.
>>
>
> And if the mobile system goes down? It did here in East Anglia, during
> Storm Bebet.
> If the VoIP is down, then the ethernet is probably down too?

We should be clear about the nature of these failures.

Ordinarily, I would expect a telecare system to work using VoIP. Others
here have explained why an old design might not. This is an issue that
has been known about for at least 20 years and it behoves the telecare
providers to update their systems accordingly.

The problem of the supporting connection infrastructure is quite
separate. So if a falling tree brings down your copper landline then a
traditional telecare system will fail.

In the same way a failing broadband connection will break a telecare
system, regardless of whether that broadband is carried over a copper
pair or optical fibre; and regardless of whether the failure is caused
by a failure of the electricity supply or a failure of the copper or
fibre infrastructure.

However the broadband supplier can (in principle) determine immediately
that a customer's connection has failed and could potentially provide
this information to a telecare service provider - but as far as I know
this isn't actually done - yet. I don't think a telecare system using
the traditional copper pair could provide an equivalent immediate alert
- but some might initiate an "All OK" call at regular intervals.

So there is scope for improvement. Traditional landline-based telecare
systems usually contain an integrated backup battery - this is enough to
alert the management of a power failure so emergency provision can be made.

There is no good reason why a telecare provider could not also provide a
battery backup for an internet-based system. They could also provide a
4G or 5G backup of copper/fibre for the internet connection itself.
Given the specialisms involved they might be better advised to specify
that their customers buy a resilient connection from a suitable
supplier. Of course there will be people whose geographic location
means they can't get a mobile signal, but I might then suggest that the
customer relocates to somewhere less remote. It's all very well to be
able to call for help, but if a storm has brought down power lines then
it has probably blocked roads and washed away bridges.

So it comes down to whether one can make reasonable provision for easily
predictable problems.


--
Graham J

[Mark Carver]

On 12/02/2024 09:42, Tweed wrote:

>
> Once power and comms goes down for a prolonged period the vulnerable need
> to be visited and/or moved. A fall alarm doesn’t keep someone warm and fed.
>
>

This is very true. Over the 55 years the phone line at my mother's (ex
family home) existing, there have been many failures and impairments on
the phone line owing to tree/wind damage, but far fewer problems with
electricity

[Tweed]

Ethernet (and/or data comms via the mobile network) connectivity is
interesting. That allows for continuous system functionality monitoring.
Current dial up systems give no indication to the monitoring centre that
the equipment is in working order. Once a data link is used there’s no real
need to worry about DTMF tones traversing a voip link. An immediate
indication of equipment failure/power loss allows plan B (ring relatives,
neighbours to arrange a visit etc) to be activated faster than happens at
the moment. If a fall alarm is installed a plane B and ideally plans C and
D need to be in place. Simply obsessing about the power cut scenario
ignores many other failure modes.

[Davey]

On Mon, 12 Feb 2024 09:34:06 +0000

Something that would maintain the same connection reliability that is
there at the moment, which is pretty robust. I do not remember any time
since I moved here in late 2010 that the landline here was unavailable,
but the mobile signal, which is normally at least adequate if not
brilliant, has been interrupted. The landline also means that my ADSL
connection is also reliable, and it is fast enough for my modest needs.
It is all underground, except for the last few connection to the house.

It is not for me to devise a replacement system, that is for the
folks who are removing the old system.

An aside:
My only personal experience of a telecare system in use was
when my neighbour's house suffered a burst pipe one winter, when he was
away for Christmas. The system somehow realised that there was some
kind of problem (I have no idea how, maybe it suffered a short
circuit, maybe it had some automatic trigger mechanism on the
hand-held button) and called the first listed contact number, a
different neighbour, who answered the call and heard the sound of
running water. On entering the house, he discovered water literally
running down the walls and the stairway from the broken pipework in the
loft.

My wife once had an emergency wrist watch, but it relied on a mobile
signal, so would have been useless during Storm Babet. She had died by
then, however, so that was not a problem.

--
Davey.

[Richmond]

My experience of power cuts is that they are localised, maybe less than
100 properties. At those times mobile internet has always worked,
because there are plenty of masts. And I now see that 2G masts can be
quite a long way away.

My Aged P's experience of power cuts is that they have been caused by
problems in the house.

If you have a power cut the lights go out, and under those circumstances
I think a fall is more likely than usual. So, when there is a power cut,
that's when you really want your fall alarm to work. Putting an
emergency phone somewhere where you have to find it in the dark (which
is what VM does) is not a substitute.

The new fall alarm which I have installed can use any of the four mobile
networks.

[Tweed]

Thought should be given to emergency lighting. Public buildings and certain
rental properties have this requirement, but curiously it’s never mentioned
in the domestic context, especially where the elderly are present. For a
few quid you can buy battery operated PIR LED lights. I’ve got a couple to
find my way to the loo at night without turning on the full mains light.
Best to prevent the need to operate the fall alarm in the first place.

[Mark Carver]

I've just asked my mum's supplier about their box, it too will use any
available mobile network to connect, and if it can't will communicate
via the Ethernet connection.

Her existing system has an inertia device she wears on her wrist
(waterproof etc) If it detects a fall, it gets the base unit to phone
the control centre.
They answer, and their voice comes out of the base unit at high volume.
If they hear a response, and it's a false alarm, that's that.

If they don't hear a response, or the response is 'trouble', then they
call me. If I don't answer, or I tell them I'm in New Zealand etc, then
they call the emergency services.

My mum triggers the alarm erroneously 5 to 6 times a year, I get a call,
sometimes I've phoned her back, and she's been blissfully unaware of
anything wrong, other times I've phoned, had no answer, gone round
there and she's been blissfully unaware. (She's very deaf, and not very
disciplined with her hearing aids)

Three times in as many years, I've gone round there, and she's on the
floor, so it's 999 and she's been carted off to A&E.

She has daily carers, they test the system every month.

[Rupert Moss-Eccardt]

ABC - alarm by carrier. Using a signal outside the voice band to send
data.

The BT version is known as Redcare in many places. People will say
that doesn't work with broadband. Older versions didn't but newer
versions do.

[Graham J]

Redcare was often hard-wired, thereby preventing the end-user from
fitting the microfilter needed to support ADSL.

--
Graham J

[Richmond]

If there is room outside the voice band, why didn't they ever use it to
send voice? They could have electronically (but not digitally) spread
shifted the voice up during transmission and then back down again to the
speaker, achieving a wider frequency range.

[Tim+]

Probably a lot easier to do it digitally (which is what they did).

https://en.wikipedia.org/wiki/Digital_access_carrier_system

Tim

--
Please don't feed the trolls

[Richmond]

But before computers were invented, why didn't they use an electronic
system? I think it would have been possible. If you can amplify a sound
like Dolby, can you not also change its pitch?

[Graham J]

Richmond wrote:

[snip]

>
> If there is room outside the voice band, why didn't they ever use it to
> send voice? They could have electronically (but not digitally) spread
> shifted the voice up during transmission and then back down again to the
> speaker, achieving a wider frequency range.

The copper pair is really 1880's technology so being able to use a wider
spectrum wasn't really an option.

But ISDN did exactly that, giving two 56kbits/sec channels for digitised
voice, plus a 16kbits/sec channel for signalling. Broadcasters would
bond the channels to achieve higher quality audio but nobody wanted
better audio for ordinary telecomms purposes. Even, now the audio
quality of many mobile connections is absolute crap, just listen to any
broadcast telephone interviews!

In Germany many subscribers had ISDN starting in about 1990. It wasn't
really available in the UK until about 10 years later - really a missed
opportunity.



--
Graham J

[Tim+]

Richmond <dnom...@gmx.com> wrote:

> Tim+ <timdo...@yahoo.co.youkay> writes:
>

>>> If there is room outside the voice band, why didn't they ever use it
>>> to send voice? They could have electronically (but not digitally)
>>> spread shifted the voice up during transmission and then back down
>>> again to the speaker, achieving a wider frequency range.
>>>
>>
>> Probably a lot easier to do it digitally (which is what they did).
>>
>> https://en.wikipedia.org/wiki/Digital_access_carrier_system
>>
>> Tim
>
> But before computers were invented, why didn't they use an electronic
> system? I think it would have been possible. If you can amplify a sound
> like Dolby, can you not also change its pitch?
>

I know very little about electronics but I would guess that analogue
amplification is a doodle compared to analogue frequency shifting.

I’m sure someone who actually knows will be along to explain it.

[Theo]

Richmond <dnom...@gmx.com> wrote:
> If there is room outside the voice band, why didn't they ever use it to
> send voice? They could have electronically (but not digitally) spread
> shifted the voice up during transmission and then back down again to the
> speaker, achieving a wider frequency range.

In analogue days they limited the frequency range of a call in order to
use frequency shifting to fit multiple calls down a (coaxial) trunk cable:
https://en.wikipedia.org/wiki/Frequency-division_multiplexing
(or microwave/satellite/etc)

With digital trunks, the frequency limit restricts the bit rate needed to
transmit the call digitally. Means you can fit more calls in the same
bandwidth.

Given the filtering at the exchange, it leaves the local loop free for other
things at non-voice frequencies. Most commonly that's DSL these days. But
those high frequencies don't make it beyond the exchange, so it's only
useful for strictly local connections.

Theo

[Rupert Moss-Eccardt]

There used to be all sort of filters and gubbins in the old core to
keep things tidy, which would block things like broadband and ABC. ABC
broke out at the exchange to its own backhaul to central stations and
the broadband overlay network did the same for broadband when it came
in.

[Mike Humphrey]

They did. I had to go digging for the name, but it seems it was WB900.
There's a description and a lot of photos here -
https://telephonesuk.org.uk/shared-service/ - and if you Google "wb900"
there's a few other pages about it.

The general term is "pair gain", and the Wikipedia article mentions
systems allowing up to 60 lines on a single pair. I don't think any more
than 2 was used in the UK - though as ISDN30 can provide 30 voice channels
over a copper pair it would have been possible to base something on that.

Mike

[Abandoned Trolley]

> The general term is "pair gain", and the Wikipedia article mentions
> systems allowing up to 60 lines on a single pair. I don't think any more
> than 2 was used in the UK - though as ISDN30 can provide 30 voice channels
> over a copper pair it would have been possible to base something on that.
>
> Mike

60 lines on a single pair ? - sounds a bit like 5 groups in a supergroup ?

I think it predates ISDN 30 though


Unless you're thinking of "DCME" ? (which also predates ISDN 30)

[Mike Humphrey]

On Tue, 20 Feb 2024 14:50:55 +0000, Abandoned Trolley wrote:
>> The general term is "pair gain", and the Wikipedia article mentions
>> systems allowing up to 60 lines on a single pair. I don't think any
>> more than 2 was used in the UK - though as ISDN30 can provide 30 voice
>> channels over a copper pair it would have been possible to base
>> something on that.
>

> 60 lines on a single pair ? - sounds a bit like 5 groups in a
> supergroup?
>
> I think it predates ISDN 30 though
>
> Unless you're thinking of "DCME" ? (which also predates ISDN 30)

I don't know, just going on the Wikpedia article, which is quite light on
detail. As far as I'm aware, pair gain in the UK never went beyond 2-
channel (WB900, and then DACS).
I just mentioned ISDN30 over copper as proof that 30 voice channels could
fit into a single copper pair.

Mike

[Abandoned Trolley]

OK

Before ISDN30 became a "thing" in this country, the Post Office rolled
out some 24 channel PCM kit which was made by Plessey in a factory in
Upminster (remember factories ?) - I was a TT(A) at the time.

One of the reasons for the slow rollout of "new" technology in this
country seems the almost interminable installation contracts - just for
one example, as a contractor in 1991 I had an interview for a job with a
well known telecoms supplier, which involved finishing up a contract
with BT to complete their 140 Mbit PDH network - using coax !

[Rupert Moss-Eccardt]

I used Pairgain modems on some private wires I had to provide a 2Mb/s
network in Cambridge.

Not all University buildings were on the Grant Backbone Network (GBN)
but many were on the CUDN (Cambridge University Data Network).

So, for example, to get from the University administration buildings
(aka The Old Schools) to, for example, the Assistant Staff Office, I
connected a Pairgain pair (effectively ADSL so needed two) to the
"Engineering Order Wire" at the GBN egress in the Old Schools which
took me to the Computing Service. From there I patched across to the
CUDN two pair that went to the RJE hut in the Assistant Staff garden
and then up into the main building to the other pair of Pairgains.

They weren't BABT approved but that didn't matter as this was all
private wire.