[[email protected]]

A recent article in the electronic newsletter for FIdonet
states the voice over internet protocol will supplant plain
old telephone service in most of the world soon. MIght be
true. IN either case, I have some problems with that.
HEre's my response. COmments anyone?

IS there a future for pots? another viewpoint


I read the article in 2536 with some interest, as I"m one of those with pots
connectivity only. Although many are sold
on voice over internet protocols I'm not convinced, and hope that our
telecommunications companies here in North America
avoid the switch for a variety of reasons. I'll list them
in order of most important to the public down to my personal reasons.

first and foremost, VoIp requires more infrastructure to get the audio from
point a to point b. In most advanced
societies services such as 9-1-1 for emergency response
dispatch are readily available, and we've grown to depend on them. with the
simple pots technology we've become used to
handling it can remain reliable even after weather events
such as storms do major damage. The reason is that most of
the landline cabling is buried underground in many parts of
the U.s. Combine this with reliable battery backup for
switching facilities and you've got a robust system that
will still continue to handle calls even after the trees
have blown down and the power grid is knocked out. I'm such a firm believer in
maintaining simple pots capability that
I've convinced many elderly family members not to give up
their pots lines for cell phones only. while arguing the
point with them I tell them not to abandon their plain old
telephone handset with keypad or dial in favor of the
wireless systems and others that also require power from the grid. IT just may
save a life.

voIp on the other hand requires more infrastructure to
remain working. You've got to convert that audio to data
packets, and route them. THe more complex you make a system the more vulnerable
it is to failure either due to tampering or just accident. Many times when
operating an emergency
radio communications facility I've hooked up a small battery to a field
telephone like armies use and strung some wire to the folks
I'm providing
communications for. THis way, they
can communicate directly with me without having to send a courier from one place
to another. I can also phone patch them
directly onto the radio if the need should arise. it's
simple, a low voltage power source, some wire and two simple handsets. THe
regular landline telephone you find in your home isn't
much more complicated than that. OTher than the dual tone
generator which is driven by the same power source as the
rest of the system it's just a simple audio connection. The twisted pair
balanced line does fairly good at keeping
interference out and delivering audio to each end of the
connection reliably.

This simple system is easier to troubleshoot, easier to
restore to service after the large scale outage. STring
another line to patch around the trouble spot, add more DC
power; if all else fails go back to electromechanical
relays. Easy to use; easy to fix; high degree of
reliability. What's not to like?

Many claim not to hear a difference in audio quality for
voice calls routed via VoIp or via regular pots connections. Wish I could say
that. I've used VoIp modes communicating
over ham radio, and communicated many times with people
using such voIp phone services. I'm appalled that we would
consider the poor quality audio we get from these digital
cell phones and VoIp connections as acceptable. Telephone
audio has become poorer just in my fifty plus years on this
planet, and it should be getting better. I seem to spend
more time with these newfangled digital packet switched
audio connections saying "what was that again?" than I
should be.


THen again, there are plenty of folks in the developed world still on dial-up
internet connections. There are still many
places here in the Americas where you're too far away from a switch to get dsl
reliably. IN many of them there is not
cable TV service.

IN short, those of us with a few clues should be telling the telecomm service
providers that we expect reliability and
better sound. THe first time your packet switched cool voIp system causes
Grandma not to get emergency services MR.
TElecomm CEO I hope they drag you through the courts for the rest of your
natural life.

Just $0.02 worth from out here in the trenches.




Richard webb,
replace anything before at with elspider

[Don Pearce]

wrote:
A recent article in the electronic newsletter for FIdonet
states the voice over internet protocol will supplant plain
old telephone service in most of the world soon. MIght be
true. IN either case, I have some problems with that.
HEre's my response. COmments anyone?

IS there a future for pots? another viewpoint


I read the article in 2536 with some interest, as I"m one of those with pots
connectivity only. Although many are sold
on voice over internet protocols I'm not convinced, and hope that our
telecommunications companies here in North America
avoid the switch for a variety of reasons. I'll list them
in order of most important to the public down to my personal reasons.

first and foremost, VoIp requires more infrastructure to get the audio from
point a to point b. In most advanced
societies services such as 9-1-1 for emergency response
dispatch are readily available, and we've grown to depend on them. with the
simple pots technology we've become used to
handling it can remain reliable even after weather events
such as storms do major damage. The reason is that most of
the landline cabling is buried underground in many parts of
the U.s. Combine this with reliable battery backup for
switching facilities and you've got a robust system that
will still continue to handle calls even after the trees
have blown down and the power grid is knocked out. I'm such a firm believer in
maintaining simple pots capability that
I've convinced many elderly family members not to give up
their pots lines for cell phones only. while arguing the
point with them I tell them not to abandon their plain old
telephone handset with keypad or dial in favor of the
wireless systems and others that also require power from the grid. IT just may
save a life.

voIp on the other hand requires more infrastructure to
remain working. You've got to convert that audio to data
packets, and route them. THe more complex you make a system the more vulnerable
it is to failure either due to tampering or just accident. Many times when
operating an emergency
radio communications facility I've hooked up a small battery to a field
telephone like armies use and strung some wire to the folks
I'm providing
communications for. THis way, they
can communicate directly with me without having to send a courier from one place
to another. I can also phone patch them
directly onto the radio if the need should arise. it's
simple, a low voltage power source, some wire and two simple handsets. THe
regular landline telephone you find in your home isn't
much more complicated than that. OTher than the dual tone
generator which is driven by the same power source as the
rest of the system it's just a simple audio connection. The twisted pair
balanced line does fairly good at keeping
interference out and delivering audio to each end of the
connection reliably.

This simple system is easier to troubleshoot, easier to
restore to service after the large scale outage. STring
another line to patch around the trouble spot, add more DC
power; if all else fails go back to electromechanical
relays. Easy to use; easy to fix; high degree of
reliability. What's not to like?

Many claim not to hear a difference in audio quality for
voice calls routed via VoIp or via regular pots connections. Wish I could say
that. I've used VoIp modes communicating
over ham radio, and communicated many times with people
using such voIp phone services. I'm appalled that we would
consider the poor quality audio we get from these digital
cell phones and VoIp connections as acceptable. Telephone
audio has become poorer just in my fifty plus years on this
planet, and it should be getting better. I seem to spend
more time with these newfangled digital packet switched
audio connections saying "what was that again?" than I
should be.


THen again, there are plenty of folks in the developed world still on dial-up
internet connections. There are still many
places here in the Americas where you're too far away from a switch to get dsl
reliably. IN many of them there is not
cable TV service.

IN short, those of us with a few clues should be telling the telecomm service
providers that we expect reliability and
better sound. THe first time your packet switched cool voIp system causes
Grandma not to get emergency services MR.
TElecomm CEO I hope they drag you through the courts for the rest of your
natural life.

Just $0.02 worth from out here in the trenches.




Richard webb,
replace anything before at with elspider



The UK telephone network backbone is VoIP, and I have experienced no
problems. I know of no plans to attempt to take VoIP any further than
the switch - there is no need while everyone has their own copper pair
to the house. Of course services like Skype run on systems without GOS
specifications, intended for non-realtime data transfer so it is no
surprise that they perform badly.

d

[Geoff]

Don Pearce wrote:
The UK telephone network backbone is VoIP, and I have experienced no
problems. I know of no plans to attempt to take VoIP any further than
the switch - there is no need while everyone has their own copper pair
to the house. Of course services like Skype run on systems without GOS
specifications, intended for non-realtime data transfer so it is no
surprise that they perform badly.

VoIP arther than dedicated PCM ? You sure ?

At a sunscriber-end solution I've never seen (heard) anything but
unsatisfactory with VoIP over a sustained period.

geoff

[Jay Ts]

geoff wrote:
Don Pearce wrote:
The UK telephone network backbone is VoIP, and I have experienced no
problems. I know of no plans to attempt to take VoIP any further than
the switch - there is no need while everyone has their own copper pair
to the house. Of course services like Skype run on systems without GOS
specifications, intended for non-realtime data transfer so it is no
surprise that they perform badly.

VoIP arther than dedicated PCM ? You sure ?

A couple of details may need attention. First, the old-style
of digital long distance was packet-switched, highly-compressed
to (IIRC) 6kHz and 4 bits. That was transferred using a digital
system that guaranteed 120 ms. of delay.

Although Skype uses VoIP, it is also possible to run a VoIP
system that is independent of the Internet, making it possible
to implement a higher quality of service. It is essentially
like packet switching without having its QoS.

With sufficiently high-bandwidth networking hardware, it
is possible to implement telephone connections with VoIP
that are (arguably?) about as good as the older packet
switching network. And it is a lot cheaper.

[Corrections welcome -- I was treading thin ice through all of
that, going on my memory of a talk given by the head of the
Cisco's Phoenix sales office several years ago, when they'd
just come out with their new VoIP product, and were trying to
sell it to corporations as a cheaper replacement for international
T1 lines.]

Jay Ts
--
To contact me, use this web page:
http://www.jayts.com/contact.php

[Richard Crowley]

"Jay Ts" wrote ...
A couple of details may need attention. First, the old-style
of digital long distance was packet-switched, highly-compressed
to (IIRC) 6kHz and 4 bits. That was transferred using a digital
system that guaranteed 120 ms. of delay.

Although Skype uses VoIP, it is also possible to run a VoIP
system that is independent of the Internet, making it possible
to implement a higher quality of service. It is essentially
like packet switching without having its QoS.

With sufficiently high-bandwidth networking hardware, it
is possible to implement telephone connections with VoIP
that are (arguably?) about as good as the older packet
switching network. And it is a lot cheaper.

[Corrections welcome -- I was treading thin ice through all of
that, going on my memory of a talk given by the head of the
Cisco's Phoenix sales office several years ago, when they'd
just come out with their new VoIP product, and were trying to
sell it to corporations as a cheaper replacement for international
T1 lines.]

Large corps (such as my employer) have been doing inter-site
VOIP for many years. Most systems like that overflow to the
public switched network when the VOIP is full, but none of
us have ever been able to detect any difference in QOS. Of
course they use very heavy-duty hardware encryption :-)

[Geoff]

Jay Ts wrote:
geoff wrote:
Don Pearce wrote:
The UK telephone network backbone is VoIP, and I have experienced no
problems. I know of no plans to attempt to take VoIP any further
than the switch - there is no need while everyone has their own
copper pair to the house. Of course services like Skype run on
systems without GOS specifications, intended for non-realtime data
transfer so it is no surprise that they perform badly.

VoIP arther than dedicated PCM ? You sure ?

A couple of details may need attention. First, the old-style
of digital long distance was packet-switched, highly-compressed
to (IIRC) 6kHz and 4 bits. That was transferred using a digital
system that guaranteed 120 ms. of delay.

4KHz bw and 8 bits. Mind you, that was me in dedicated hardware pre-packet
switching PCM days, so it might have changed in the meantime (!)..


With sufficiently high-bandwidth networking hardware, it
is possible to implement telephone connections with VoIP
that are (arguably?) about as good as the older packet
switching network. And it is a lot cheaper.

That the 'in theory' bit. In practice the latency seems variable even on
Gigabit Ethernet.

Wass the packet-switching phase of PCM telephone trunk transmission on a
'per call' or a 'per system' basis ?

geoff

[Scott Dorsey]

Jay Ts wrote:

A couple of details may need attention. First, the old-style
of digital long distance was packet-switched, highly-compressed
to (IIRC) 6kHz and 4 bits. That was transferred using a digital
system that guaranteed 120 ms. of delay.

It was not packet-switched. It was circuit switched, so each channel
had guaranteed bandwidth.

I don't know where that delay spec came from... I have had T-1 circuits
especially over satellites that had way more than 120 ms of delay. The
good news is that the delay was constant.

Although Skype uses VoIP, it is also possible to run a VoIP
system that is independent of the Internet, making it possible
to implement a higher quality of service. It is essentially
like packet switching without having its QoS.

It IS packet switching. Normally systems like this use IP packet switching
with QoS management on top of IP. This does not give you the guaranteed
bandwidth of a circuit-switched network, but you can engineer a system
that works well most of the time by throwing bandwidth at it and keeping
circuit utilization down. Bandwidth is cheap (and the bandwidth utilization
doing this is still a lot higher than with a circuit-switched network).

With sufficiently high-bandwidth networking hardware, it
is possible to implement telephone connections with VoIP
that are (arguably?) about as good as the older packet
switching network. And it is a lot cheaper.

I wouldn't say a LOT cheaper, because it's expensive to do it right. But
it's possible to do it well.

[Corrections welcome -- I was treading thin ice through all of
that, going on my memory of a talk given by the head of the
Cisco's Phoenix sales office several years ago, when they'd
just come out with their new VoIP product, and were trying to
sell it to corporations as a cheaper replacement for international
T1 lines.]

The problem is that if you do this, you still need either international
T-1 lines in order to get control over all the bandwidth in your private
network, OR you need a frame relay network where the telco gives you
guarantees about quality of service. Admittedly you can get a good bit
more channels over a T-1 running VoIP than you can running straight voice
with SS7, even with good quality. You can get a huge amount more if you
can put up with lousy quality.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[John Cafarella]

On Sep 9, 12:39*pm, (Scott Dorsey) wrote:

The problem is that if you do this, you still need either international
T-1 lines in order to get control over all the bandwidth in your private
network, OR you need a frame relay network where the telco gives you
guarantees about quality of service. *Admittedly you can get a good bit
more channels over a T-1 running VoIP than you can running straight voice
with SS7, even with good quality. *You can get a huge amount more if you
can put up with lousy quality.
--scott
--
"C'est un Nagra. *C'est suisse, et tres, tres precis."

You're right on pretty much every count in your post Scott.
I'm an operations manager at a decent sized telco, we indeed do use
VOIP as a good portion of our backbone, but this is on a carefully
managed private network, and throwing large amounts of bandwidth is
key to sucess, along with very careful attention to routing and backup
routing. Voice quality is indistinguishable from traditional PCM.

Instead of using frame relay with guarantees of service, the more
usual implementation these days is an end to end managed IP based
Virtual Private Network (IVPN), with well designed QOS.

An IP based backbone can deliver perfectly good quality, but it DOES
have to be carefully managed. However the same is true for a TDM
based network, it's just that the technology and techniques are more
mature, and therefore trivial. Low cost players don't always have the
knowledge and training to run a network well.

John

[Kevin T]

On Sep 8, 10:39*pm, (Scott Dorsey) wrote:
Jay Ts wrote:



A couple of details may need attention. First, the old-style
of digital long distance was packet-switched, highly-compressed
to (IIRC) 6kHz and 4 bits. That was transferred using a digital
system that guaranteed 120 ms. of delay.

It was not packet-switched. *It was circuit switched, so each channel
had guaranteed bandwidth.

I don't know where that delay spec came from... I have had T-1 circuits
especially over satellites that had way more than 120 ms of delay. *The
good news is that the delay was constant.

Although Skype uses VoIP, it is also possible to run a VoIP
system that is independent of the Internet, making it possible
to implement a higher quality of service. *It is essentially
like packet switching without having its QoS.

It IS packet switching. *Normally systems like this use IP packet switching
with QoS management on top of IP. *This does not give you the guaranteed
bandwidth of a circuit-switched network, but you can engineer a system
that works well most of the time by throwing bandwidth at it and keeping
circuit utilization down. *Bandwidth is cheap (and the bandwidth utilization
doing this is still a lot higher than with a circuit-switched network).

With sufficiently high-bandwidth networking hardware, it
is possible to implement telephone connections with VoIP
that are (arguably?) about as good as the older packet
switching network. And it is a lot cheaper.

I wouldn't say a LOT cheaper, because it's expensive to do it right. *But
it's possible to do it well.

[Corrections welcome -- I was treading thin ice through all of
that, going on my memory of a talk given by the head of the
Cisco's Phoenix sales office several years ago, when they'd
just come out with their new VoIP product, and were trying to
sell it to corporations as a cheaper replacement for international
T1 lines.]

The problem is that if you do this, you still need either international
T-1 lines in order to get control over all the bandwidth in your private
network, OR you need a frame relay network where the telco gives you
guarantees about quality of service. *Admittedly you can get a good bit
more channels over a T-1 running VoIP than you can running straight voice
with SS7, even with good quality. *You can get a huge amount more if you
can put up with lousy quality.
--scott
--
"C'est un Nagra. *C'est suisse, et tres, tres precis."

Scott

Are there any stats on current average MOS or PESQ etc. for most VOIP
customers. I believe the old AT&T POTS had ~ 4.0 to 4.5 MOS benchmark.
Lucent wireless was very happy with ~ 3.6.Land to Mobile. How far have
the Vonage type providers degraded the Bell Labs audio quality ? How
far can they push it until Joe average looks elsewere ?

[Meindert Sprang]

wrote in message
.. .
A recent article in the electronic newsletter for FIdonet
states the voice over internet protocol will supplant plain
old telephone service in most of the world soon. MIght be
true. IN either case, I have some problems with that.
HEre's my response. COmments anyone?

IS there a future for pots? another viewpoint


Just for your information: POTS was already digital between switchboard
before the internet became common place. Compared to the "ancient" digital
connections, VOIP is just another protocol over the same physical
connection, being copper or glass. Since eons, optical fibre is used for
long distance and high capacity telephone links. Again, long before the
internet became what it is now. And all those fibre links needed more
electronics than the plain old copper twisted pair and a lead-acid battery.

Meindert

[[email protected]]

On 2008-09-09 said:
Just for your information: POTS was already digital between
switchboard before the internet became common place. Compared to
the "ancient" digital connections, VOIP is just another protocol
over the same physical connection, being copper or glass. Since
eons, optical fibre is used for long distance and high capacity
telephone links. Again, long before the internet became what it is
now. And all those fibre links needed more electronics than the
plain old copper twisted pair and a lead-acid battery.
AGreed, but the end user's connection is on said copper,
backed up by generator power or said lead acid battery
grin. NOte my comment about the hospital being without
its phone service, but a pots line across the street still
being available during the hurricane. From that phone on a
desk to the switch was copper, with backup power of some
type. Long distance was an iffy proposition on that line,
but you could still reliably dial across town, and
presumably reach the local 9-1-1 answering point.

REgards,




Richard webb,
replace anything before at with elspider