Phase modulation detection/NIST plan
On Wed, Jul 11, 2012 at 03:48:52PM -0400, paul swed wrote:
David
Read your comments and have been traveling. So finally a chance to email.
I read the document also and walked away with what I shared.
In your reading would you believe the following.
Its an absolute phase and that when it switches to 0 there is 1 transition
at the beginning of the second to 180 degrees staying that way to the next
bit or flipping again to 0 degrees if its a 1 at the 1 sec tic???
What I mean by absolute phase is that a 1 is always 180 degrees
and a zero always 0 degrees. In your example this would imply that the
two ones in a row would result in two seconds of 180 degree phase
without a flip after the first 1.
The document is confusing, but the best I can do with its
language is to conclude they are talking about absolute phase. Normally
when one talks about baseband waveforms one is referring to absolute I
and Q components relative to an unchanging carrier phase, not relative I
and Q with respect to the last bit phase. So I take their language to
mean a zero is 0 degrees and a 1 180 degrees relative to an unchanging
carrier.
Differential encoding is the opposite, a 1 is always the
opposite phase from the last bit, a zero always the same phase as the
last bit (or sometimes the inverse where a zero is the transition and a
one is not).
Is there a way to sense from the document that there is a bias towards 0
lets say.
Differential encoding tends to have little "DC" component or
bias toward either one or zero or one phase or the other, absolute
encoding does if the data it encodes does.
--
"An empty zombie mind with a forlorn barely readable weatherbeaten
'For Rent' sign still vainly flapping outside on the weed encrusted pole - in
celebration of what could have been, but wasn't and is not to be now either."
On 07/14/2012 01:49 AM, David I. Emery wrote:
On Wed, Jul 11, 2012 at 03:48:52PM -0400, paul swed wrote:
David
Read your comments and have been traveling. So finally a chance to email.
I read the document also and walked away with what I shared.
In your reading would you believe the following.
Its an absolute phase and that when it switches to 0 there is 1 transition
at the beginning of the second to 180 degrees staying that way to the next
bit or flipping again to 0 degrees if its a 1 at the 1 sec tic???
What I mean by absolute phase is that a 1 is always 180 degrees
and a zero always 0 degrees. In your example this would imply that the
two ones in a row would result in two seconds of 180 degree phase
without a flip after the first 1.
The document is confusing, but the best I can do with its
language is to conclude they are talking about absolute phase. Normally
when one talks about baseband waveforms one is referring to absolute I
and Q components relative to an unchanging carrier phase, not relative I
and Q with respect to the last bit phase. So I take their language to
mean a zero is 0 degrees and a 1 180 degrees relative to an unchanging
carrier.
I think the PTTI article isn't as much documentation as presentation of
general principle, showing details more as to present how it can be
done, but not necessarily guarantee it will be done that way. Knowing
the synchronisation sequence, polarity should not be ambiguous. Also
note that other data such as hours would be known from the AM signal, so
we can reverse engineer it. A receiver knowing this sequence will either
bootstrap from the AM or attempt straight lock. It's not too hard to
build a maximum likelihood receiver for it.
Cheers,
Magnus
On Sat, Jul 14, 2012 at 02:38:34AM +0200, Magnus Danielson wrote:
I think the PTTI article isn't as much documentation as presentation of
general principle, showing details more as to present how it can be
done, but not necessarily guarantee it will be done that way. Knowing
the synchronisation sequence, polarity should not be ambiguous. Also
note that other data such as hours would be known from the AM signal, so
we can reverse engineer it. A receiver knowing this sequence will either
bootstrap from the AM or attempt straight lock. It's not too hard to
build a maximum likelihood receiver for it.
I read the article as not a definitive specification frozen in
stone, but as a complete and relatively fully specified proposed design
with perhaps some details subject to adjustment or revision.
The question of absolute versus differential phase shift keying
is, of course, rather fundamental to being able to decode the signal at
one level but at another not terribly central to the core of the design
for a coding and modulation scheme that works at much lower C/N levels
than the AM version did while preserving the legacy AM and its coding
for existing hardware.
SOME place in the design of a differentially coded signal there
has to be a decision whether or not to structure the data encoding so
some specific bit (or more properly symbol) in each frame (or at least
some known frames relative to the time of day) (in this case I mean 1
minute long TOD frame) is of a known absolute reference phase.
If this is done than it becomes possible in a reasonable time to
determine an absolute 60 KHz carrier phase after a fade, if it is not
done and every single bit of data is not absolutely predictable (the
current TOD coding would be absolutely predictable given knowledge of
the time and date and of leap seconds and DST settings, but they make
clear future extensions would probably not have this property as
additional messages are added including emergency messages and the like
which are never predictable) there is no way to reliably decide after a
fade which phase is which as this depends on knowing the number of ones
and zeros in all the frames transmitted since one last saw the signal,
something that is in the general case impossible if the signal has faded
and the bits were not observed.
An absolute encoding has no ambiguity - if one knows the time of
day within a second one knows the transmitted phase except for during
bits that might vary with unknown data (eg emergency messages,
extensions to the standard and newly changed DST and leap second
settings and FEC bits based on them) and MOST bits are always known
phases, especially of course the sync code words. So even with
terribly poor C/N one should be able to relatively quickly resolve the
phase ambiguity after a period of signal loss... and in many cases if
one still has a good idea of the time, within a couple of seconds
(symbols) of signal reacquisition.
On another point, I am not of the school that providing much
better weak signal performance for simple, low power, and cheap LF time
of day clocks using WWVB is somehow a minor improvement that primarily
benefits China because they make most cheap self setting "atomic"
clocks. There are innumerable applications for low cost low power
human level 1 second accurate time of day in modern electronic systems -
examples are traffic lights and school crossing signs and water
sprinklers and street lights and other outdoor lighting and many
others... these systems are not normally network connected and there is
no current wide area technology short of power hungry GPS with its weak
signals and relatively high cost and difficult reception from many
locations to do this.
And with minimal effort to ensure compatibility, there should be
no conflict with use of the same carrier signal as a frequency reference
too... the problem of several decades old antique time and frequency
gear being incompatible seems very minor, and of course we have already
discussed ways to handle this if needed.
And as long as the existing frequency reference use of the
carrier continues to work as a backup to GPS with modern updated gear
that capability hasn't been lost - except maybe if your particular
variety of tin foil hat requires vacuum tube VLF reference gear because
of EMP fears or something similar.
I think the new WWVB proposal seems sensible and a reasonable
design...that should serve the public well.
--
Dave Emery N1PRE/AE, die@dieconsulting.com DIE Consulting, Weston, Mass 02493
"An empty zombie mind with a forlorn barely readable weatherbeaten
'For Rent' sign still vainly flapping outside on the weed encrusted pole - in
celebration of what could have been, but wasn't and is not to be now either."
I received the following message from John Lowe at NIST.
I thought it might be of interest to you.
-----Original Message-----
From: John Lowe [mailto:lowe@boulder.nist.gov]
Sent: Monday, July 09, 2012 12:55 PM
To: Ron Ward
Subject: Re: Phase-locking 60 kHz timing and frequency standard
receivers
We are changing the format to improve our reception capability.
Frequency standard comparisons will still be possible with new or
modified equipment.
On 7/5/2012 9:39 AM, Ron Ward wrote:
Hi John:
Why is WWVB changing to BPSK? For cheaper clocks?
What about frequency standard comparisons with WWVB?
How am I going to monitor GPS and ensure that it's working correctly and
not being played with by DOD?
What am I going to do with our phase-locking 60 kHz timing and frequency
standard receivers?
Why did you not make the new data format backward compatible with
existing phase-locking 60 kHz timing and frequency standard receivers,
like they did with black and white to color TV transition in the 1950's?
What about +- 45 degree modulation?
We use to have LORAN, OMEGA, and WWVB.
Thanks,
Ron
-----Original Message-----
From: time-nuts-bounces@febo.com [mailto:time-nuts-bounces@febo.com] On
Behalf Of David I. Emery
Sent: Friday, July 13, 2012 8:35 PM
To: Discussion of precise time and frequency measurement
Subject: Re: [time-nuts] Phase modulation detection/NIST plan
On Sat, Jul 14, 2012 at 02:38:34AM +0200, Magnus Danielson wrote:
I think the PTTI article isn't as much documentation as presentation
of
general principle, showing details more as to present how it can be
done, but not necessarily guarantee it will be done that way. Knowing
the synchronisation sequence, polarity should not be ambiguous. Also
note that other data such as hours would be known from the AM signal,
so
we can reverse engineer it. A receiver knowing this sequence will
either
bootstrap from the AM or attempt straight lock. It's not too hard to
build a maximum likelihood receiver for it.
I read the article as not a definitive specification frozen in
stone, but as a complete and relatively fully specified proposed design
with perhaps some details subject to adjustment or revision.
The question of absolute versus differential phase shift keying
is, of course, rather fundamental to being able to decode the signal at
one level but at another not terribly central to the core of the design
for a coding and modulation scheme that works at much lower C/N levels
than the AM version did while preserving the legacy AM and its coding
for existing hardware.
SOME place in the design of a differentially coded signal there
has to be a decision whether or not to structure the data encoding so
some specific bit (or more properly symbol) in each frame (or at least
some known frames relative to the time of day) (in this case I mean 1
minute long TOD frame) is of a known absolute reference phase.
If this is done than it becomes possible in a reasonable time to
determine an absolute 60 KHz carrier phase after a fade, if it is not
done and every single bit of data is not absolutely predictable (the
current TOD coding would be absolutely predictable given knowledge of
the time and date and of leap seconds and DST settings, but they make
clear future extensions would probably not have this property as
additional messages are added including emergency messages and the like
which are never predictable) there is no way to reliably decide after a
fade which phase is which as this depends on knowing the number of ones
and zeros in all the frames transmitted since one last saw the signal,
something that is in the general case impossible if the signal has faded
and the bits were not observed.
An absolute encoding has no ambiguity - if one knows the time of
day within a second one knows the transmitted phase except for during
bits that might vary with unknown data (eg emergency messages,
extensions to the standard and newly changed DST and leap second
settings and FEC bits based on them) and MOST bits are always known
phases, especially of course the sync code words. So even with
terribly poor C/N one should be able to relatively quickly resolve the
phase ambiguity after a period of signal loss... and in many cases if
one still has a good idea of the time, within a couple of seconds
(symbols) of signal reacquisition.
On another point, I am not of the school that providing much
better weak signal performance for simple, low power, and cheap LF time
of day clocks using WWVB is somehow a minor improvement that primarily
benefits China because they make most cheap self setting "atomic"
clocks. There are innumerable applications for low cost low power
human level 1 second accurate time of day in modern electronic systems -
examples are traffic lights and school crossing signs and water
sprinklers and street lights and other outdoor lighting and many
others... these systems are not normally network connected and there is
no current wide area technology short of power hungry GPS with its weak
signals and relatively high cost and difficult reception from many
locations to do this.
And with minimal effort to ensure compatibility, there should be
no conflict with use of the same carrier signal as a frequency reference
too... the problem of several decades old antique time and frequency
gear being incompatible seems very minor, and of course we have already
discussed ways to handle this if needed.
And as long as the existing frequency reference use of the
carrier continues to work as a backup to GPS with modern updated gear
that capability hasn't been lost - except maybe if your particular
variety of tin foil hat requires vacuum tube VLF reference gear because
of EMP fears or something similar.
I think the new WWVB proposal seems sensible and a reasonable
design...that should serve the public well.
--
Dave Emery N1PRE/AE, die@dieconsulting.com DIE Consulting, Weston,
Mass 02493
"An empty zombie mind with a forlorn barely readable weatherbeaten
'For Rent' sign still vainly flapping outside on the weed encrusted pole
- in
celebration of what could have been, but wasn't and is not to be now
either."
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.