Optical SETI and optical beacons

Hi

There is a lot of talk in the group about microwave SETI, but much
less about Optical beacon search. The SETI FAQ page is also stating
more facts about microwave than about optical.

For example, we have a very nice formula for microwave, which relates
distance, feed-power, antenna size, background noise (system temp),
bandwidth and observation time (signal integration) and such for
recognizable
signal reproduction (S/N). I still miss 'pulse-width and pulse-separation'
in
that formula for pulsed signals, but other than that it seems pretty
complete.

I'm curious to know if we have (or can create) a similar formula for
optical,
which sets some theoretical limitations of optical communication.

I'm thinking about a formula which can test questions like this :

- Is there is limit to the pulse-width ? E.g. why are we checking for
ns pulses for OSETI ? Why not ps pulses, or fs ?
- How does pulse-width relate to detectability or S/N ratio ?
- How does pulse-separation relate to detectability or S/N ratio ?
- How much info can ET put in a pulsed signal ?
- How does continuous, narrowband laser relate to pulsed laser signals ?
- How does a beacon's energy / equipment cost vary with these
variables...
- etc.

Most of the micro-wave formula can probably be re-used, but there are
certain things vastly different in optical. We need at least pulse info into
the formula,
and background noise is also different : for pulsed laser, the transmitter
defines
most of the S/N ratio, not the receiver, because it needs to blast over the
light output of the sun it is close to.

Anyone knows a good overview of this ?

And some fun questions :
- What is the dispersion of pulses through interstellar space, and what
is it through our own atmosphere (to detect these pulses from space) ?
- Does the transmitter really need a telescope ? (How narrow is a laser
beam without one)
- How do you project a 1mm laser beam onto a 1m telescope (if you need
one).
- How do you detect ps pulses ?
- What would happen if an airplane flies over a laser beacon when it's
on, or a bird,
(we must assume that ET beacon designers are concerned about their
environment too).
- Did we do a moon-bounce experiment with laser pulses yet ? Or a
mars-bounce ?
Or does a rough-surface-bounce totally destroy a nice narrow pulse ?

Thanks

Rob

Dr Kingsley, the stalwart pioneer in optical SETI, has an excellent Optical
SETI web site:
(much information is provided)

http://www.coseti.org/

Al



"Rob Dekker" wrote in message
. com...
Hi

There is a lot of talk in the group about microwave SETI, but much
less about Optical beacon search. The SETI FAQ page is also stating
more facts about microwave than about optical.

For example, we have a very nice formula for microwave, which relates
distance, feed-power, antenna size, background noise (system temp),
bandwidth and observation time (signal integration) and such for
recognizable
signal reproduction (S/N). I still miss 'pulse-width and pulse-separation'
in
that formula for pulsed signals, but other than that it seems pretty
complete.

I'm curious to know if we have (or can create) a similar formula for
optical,
which sets some theoretical limitations of optical communication.

I'm thinking about a formula which can test questions like this :

- Is there is limit to the pulse-width ? E.g. why are we checking for
ns pulses for OSETI ? Why not ps pulses, or fs ?
- How does pulse-width relate to detectability or S/N ratio ?
- How does pulse-separation relate to detectability or S/N ratio ?
- How much info can ET put in a pulsed signal ?
- How does continuous, narrowband laser relate to pulsed laser signals
?
- How does a beacon's energy / equipment cost vary with these
variables...
- etc.

Most of the micro-wave formula can probably be re-used, but there are
certain things vastly different in optical. We need at least pulse info
into
the formula,
and background noise is also different : for pulsed laser, the transmitter
defines
most of the S/N ratio, not the receiver, because it needs to blast over
the
light output of the sun it is close to.

Anyone knows a good overview of this ?

And some fun questions :
- What is the dispersion of pulses through interstellar space, and what
is it through our own atmosphere (to detect these pulses from space)
?
- Does the transmitter really need a telescope ? (How narrow is a laser
beam without one)
- How do you project a 1mm laser beam onto a 1m telescope (if you need
one).
- How do you detect ps pulses ?
- What would happen if an airplane flies over a laser beacon when it's
on, or a bird,
(we must assume that ET beacon designers are concerned about their
environment too).
- Did we do a moon-bounce experiment with laser pulses yet ? Or a
mars-bounce ?
Or does a rough-surface-bounce totally destroy a nice narrow pulse ?

Thanks

Rob

Dr Kingsley, the stalwart pioneer in optical SETI, has an excellent Optical
SETI web site:
(much information is provided)

http://www.coseti.org/

Al



"Rob Dekker" wrote in message
. com...
Hi

There is a lot of talk in the group about microwave SETI, but much
less about Optical beacon search. The SETI FAQ page is also stating
more facts about microwave than about optical.

For example, we have a very nice formula for microwave, which relates
distance, feed-power, antenna size, background noise (system temp),
bandwidth and observation time (signal integration) and such for
recognizable
signal reproduction (S/N). I still miss 'pulse-width and pulse-separation'
in
that formula for pulsed signals, but other than that it seems pretty
complete.

I'm curious to know if we have (or can create) a similar formula for
optical,
which sets some theoretical limitations of optical communication.

I'm thinking about a formula which can test questions like this :

- Is there is limit to the pulse-width ? E.g. why are we checking for
ns pulses for OSETI ? Why not ps pulses, or fs ?
- How does pulse-width relate to detectability or S/N ratio ?
- How does pulse-separation relate to detectability or S/N ratio ?
- How much info can ET put in a pulsed signal ?
- How does continuous, narrowband laser relate to pulsed laser signals
?
- How does a beacon's energy / equipment cost vary with these
variables...
- etc.

Most of the micro-wave formula can probably be re-used, but there are
certain things vastly different in optical. We need at least pulse info
into
the formula,
and background noise is also different : for pulsed laser, the transmitter
defines
most of the S/N ratio, not the receiver, because it needs to blast over
the
light output of the sun it is close to.

Anyone knows a good overview of this ?

And some fun questions :
- What is the dispersion of pulses through interstellar space, and what
is it through our own atmosphere (to detect these pulses from space)
?
- Does the transmitter really need a telescope ? (How narrow is a laser
beam without one)
- How do you project a 1mm laser beam onto a 1m telescope (if you need
one).
- How do you detect ps pulses ?
- What would happen if an airplane flies over a laser beacon when it's
on, or a bird,
(we must assume that ET beacon designers are concerned about their
environment too).
- Did we do a moon-bounce experiment with laser pulses yet ? Or a
mars-bounce ?
Or does a rough-surface-bounce totally destroy a nice narrow pulse ?

Thanks

Rob

"Rob Dekker" wrote in message .com...
Hi

There is a lot of talk in the group about microwave SETI, but much
less about Optical beacon search. The SETI FAQ page is also stating
more facts about microwave than about optical.

For example, we have a very nice formula for microwave, which relates
distance, feed-power, antenna size, background noise (system temp),
bandwidth and observation time (signal integration) and such for
recognizable
signal reproduction (S/N). I still miss 'pulse-width and pulse-separation'
in
that formula for pulsed signals, but other than that it seems pretty
complete.

I'm curious to know if we have (or can create) a similar formula for
optical,
which sets some theoretical limitations of optical communication.

I'm thinking about a formula which can test questions like this :

- Is there is limit to the pulse-width ? E.g. why are we checking for
ns pulses for OSETI ? Why not ps pulses, or fs ?
- How does pulse-width relate to detectability or S/N ratio ?
- How does pulse-separation relate to detectability or S/N ratio ?
- How much info can ET put in a pulsed signal ?
- How does continuous, narrowband laser relate to pulsed laser signals ?
- How does a beacon's energy / equipment cost vary with these
variables...
- etc.

Most of the micro-wave formula can probably be re-used, but there are
certain things vastly different in optical. We need at least pulse info into
the formula,
and background noise is also different : for pulsed laser, the transmitter
defines
most of the S/N ratio, not the receiver, because it needs to blast over the
light output of the sun it is close to.

Anyone knows a good overview of this ?

And some fun questions :
- What is the dispersion of pulses through interstellar space, and what
is it through our own atmosphere (to detect these pulses from space) ?
- Does the transmitter really need a telescope ? (How narrow is a laser
beam without one)
- How do you project a 1mm laser beam onto a 1m telescope (if you need
one).
- How do you detect ps pulses ?
- What would happen if an airplane flies over a laser beacon when it's
on, or a bird,
(we must assume that ET beacon designers are concerned about their
environment too).
- Did we do a moon-bounce experiment with laser pulses yet ? Or a
mars-bounce ?
Or does a rough-surface-bounce totally destroy a nice narrow pulse ?

Thanks

Rob

I imagine that when passing through our atmosphere the laser would
diverge by about one or two (extra) arcseconds just like starlight
does in our telescopes. Maybe it would be better to use a ruby (or gas
chamber) cylinder one meter in diameter to direct the laser rather
than a mirror. The plane would probably catch fire and/or explode and
crash; the bird would probably burn/vaporize in a flash. Wouldn't a
picosecond flash give us a photon that could be detected by CCD, CMOS,
etc? Dunno if we have circuitry that could detect two short flashes
separated by a picosecond of time. Light travels about 0.3 millimeter
in a picosecond.

"Rob Dekker" wrote in message .com...
Hi

There is a lot of talk in the group about microwave SETI, but much
less about Optical beacon search. The SETI FAQ page is also stating
more facts about microwave than about optical.

For example, we have a very nice formula for microwave, which relates
distance, feed-power, antenna size, background noise (system temp),
bandwidth and observation time (signal integration) and such for
recognizable
signal reproduction (S/N). I still miss 'pulse-width and pulse-separation'
in
that formula for pulsed signals, but other than that it seems pretty
complete.

I'm curious to know if we have (or can create) a similar formula for
optical,
which sets some theoretical limitations of optical communication.

I'm thinking about a formula which can test questions like this :

- Is there is limit to the pulse-width ? E.g. why are we checking for
ns pulses for OSETI ? Why not ps pulses, or fs ?
- How does pulse-width relate to detectability or S/N ratio ?
- How does pulse-separation relate to detectability or S/N ratio ?
- How much info can ET put in a pulsed signal ?
- How does continuous, narrowband laser relate to pulsed laser signals ?
- How does a beacon's energy / equipment cost vary with these
variables...
- etc.

Most of the micro-wave formula can probably be re-used, but there are
certain things vastly different in optical. We need at least pulse info into
the formula,
and background noise is also different : for pulsed laser, the transmitter
defines
most of the S/N ratio, not the receiver, because it needs to blast over the
light output of the sun it is close to.

Anyone knows a good overview of this ?

And some fun questions :
- What is the dispersion of pulses through interstellar space, and what
is it through our own atmosphere (to detect these pulses from space) ?
- Does the transmitter really need a telescope ? (How narrow is a laser
beam without one)
- How do you project a 1mm laser beam onto a 1m telescope (if you need
one).
- How do you detect ps pulses ?
- What would happen if an airplane flies over a laser beacon when it's
on, or a bird,
(we must assume that ET beacon designers are concerned about their
environment too).
- Did we do a moon-bounce experiment with laser pulses yet ? Or a
mars-bounce ?
Or does a rough-surface-bounce totally destroy a nice narrow pulse ?

Thanks

Rob

I imagine that when passing through our atmosphere the laser would
diverge by about one or two (extra) arcseconds just like starlight
does in our telescopes. Maybe it would be better to use a ruby (or gas
chamber) cylinder one meter in diameter to direct the laser rather
than a mirror. The plane would probably catch fire and/or explode and
crash; the bird would probably burn/vaporize in a flash. Wouldn't a
picosecond flash give us a photon that could be detected by CCD, CMOS,
etc? Dunno if we have circuitry that could detect two short flashes
separated by a picosecond of time. Light travels about 0.3 millimeter
in a picosecond.

In article ,
Rob Dekker wrote:

For example, we have a very nice formula for microwave, which relates
distance, feed-power, antenna size, background noise (system temp),
bandwidth and observation time (signal integration) and such for
recognizable

The only real difference in the formula is that the noise is predominantly
quantum shot noise, rather than thermal noise. (You can also get closer
to 100% illumination efficiency, but the tranmit and receive aperture
parts of the formulae are the same.)

signal reproduction (S/N). I still miss 'pulse-width and pulse-separation'
in
that formula for pulsed signals, but other than that it seems pretty
complete.

Note that pulse width and bandwidth are interrelated, and for pulses less
than about 10 seconds in duration one should be considered almost completely
determined by the other.

There is also a shift between coherent and non-coherent pulse detection when
the pulses are not much more frequent than about 10 per second.

The threshold signal to noise ratio would need increasing, because of the
need to search over many different periods and pulse rates (S@H already has
a high thresold for narrowband signals, because of searching over many chirp
rates).

I'm curious to know if we have (or can create) a similar formula for
optical,
which sets some theoretical limitations of optical communication.

www.coseti.org, as already mentioned, although double check formulaes; I
have reported errors in the past that make it look more favourable.

- How much info can ET put in a pulsed signal ?

Information capacity is not a big issue for SETI. The intention of SETI
is to detect existence. There may be a secondary signal that has high
information content, but the pilot signal has to be easy to detect, rather
than information rich. One proposal is to use a microwave pilot and
an optical main signal.

The most efficient use of power is, probably, to send the transmission
schedule and technical parameters as a second level pilot signal and
cycle through the destinations with the main information signal.

certain things vastly different in optical. We need at least pulse info into
the formula,

Not necessarily. One optical SETI option is to do narrowband detection
using optical superhets. I think the pulse detection case was based on
producing systems that are a relatively cheap add on to high end amateur
scopes.

and background noise is also different : for pulsed laser, the transmitter
defines
most of the S/N ratio, not the receiver, because it needs to blast over the
light output of the sun it is close to.

Not if it uses dark lines in the star's spectrum, even with CW beacons. The
main noise is quantum shot noise.

Anyone knows a good overview of this ?

- Does the transmitter really need a telescope ? (How narrow is a laser
beam without one)

Yes. Firstly, all normal radio transmitters are coherent, and they
need large antennas to create directional beams; ordinary daylight or
incandescent light is the exception to coherence, not lasers. Lasers also
need large apertures to create directional beams and they need the phase
to be flat across that aperture. For example, a semiconductor laser actually
produces a beam that diverges over many degrees, but the lenses in a typical
laser pointer form a telescope that converges it into a beam whose divergence
essentially depends on the diameter of the lens. A helium-neon laser, because
of its long cavity, will only allow full amplification for signals that
have stayed within its diameter for a long distance, but is still constrained
by that diameter in determining the lower bound on the beam width.

- How do you project a 1mm laser beam onto a 1m telescope (if you need
one).

I doubt that the sort of lasers that we are talking about would have such
a small aperture, but you would simply add a diverging lens (that's what
the eyepiece does in a simple optical telescope - it provides the whole
area of of the objective lens to the small pupil of the eye, producing
a nominally parallel beam into the eye from a point-like distant source.

Note that your 1mm beam will diverge at least about 1 in 2,000, so would,
for example, be over a 100 miles across at the moon.

- How do you detect ps pulses ?

We are certainly detecting sub-nanosecond pulses as a matter of routine
in optical fibre telephone systems, but these don't have to achieve
ultimate signal to noise ratios. I don't know what the recovery time
and rise time for a typical photmultiplier would be.

- What would happen if an airplane flies over a laser beacon when it's
on, or a bird,

The plane wouldn't be allowed to, the bird might be killed or blinded;
it depends on the power density of the beam, which will be lower for a
highly directional beam. Plane's are not allowed over Arecibo when it
is transmitting and it would be unhealthy for a bird to fly too close
to the feeds, although 30 watts per square metre in the main beam is
probably less than the exposure from a mobile phone.

- Did we do a moon-bounce experiment with laser pulses yet ? Or a
mars-bounce ?

Only with a retro-reflector installed by the Apollo astronauts.

In article ,
Rob Dekker wrote:

For example, we have a very nice formula for microwave, which relates
distance, feed-power, antenna size, background noise (system temp),
bandwidth and observation time (signal integration) and such for
recognizable

The only real difference in the formula is that the noise is predominantly
quantum shot noise, rather than thermal noise. (You can also get closer
to 100% illumination efficiency, but the tranmit and receive aperture
parts of the formulae are the same.)

signal reproduction (S/N). I still miss 'pulse-width and pulse-separation'
in
that formula for pulsed signals, but other than that it seems pretty
complete.

Note that pulse width and bandwidth are interrelated, and for pulses less
than about 10 seconds in duration one should be considered almost completely
determined by the other.

There is also a shift between coherent and non-coherent pulse detection when
the pulses are not much more frequent than about 10 per second.

The threshold signal to noise ratio would need increasing, because of the
need to search over many different periods and pulse rates (S@H already has
a high thresold for narrowband signals, because of searching over many chirp
rates).

I'm curious to know if we have (or can create) a similar formula for
optical,
which sets some theoretical limitations of optical communication.

www.coseti.org, as already mentioned, although double check formulaes; I
have reported errors in the past that make it look more favourable.

- How much info can ET put in a pulsed signal ?

Information capacity is not a big issue for SETI. The intention of SETI
is to detect existence. There may be a secondary signal that has high
information content, but the pilot signal has to be easy to detect, rather
than information rich. One proposal is to use a microwave pilot and
an optical main signal.

The most efficient use of power is, probably, to send the transmission
schedule and technical parameters as a second level pilot signal and
cycle through the destinations with the main information signal.

certain things vastly different in optical. We need at least pulse info into
the formula,

Not necessarily. One optical SETI option is to do narrowband detection
using optical superhets. I think the pulse detection case was based on
producing systems that are a relatively cheap add on to high end amateur
scopes.

and background noise is also different : for pulsed laser, the transmitter
defines
most of the S/N ratio, not the receiver, because it needs to blast over the
light output of the sun it is close to.

Not if it uses dark lines in the star's spectrum, even with CW beacons. The
main noise is quantum shot noise.

Anyone knows a good overview of this ?

- Does the transmitter really need a telescope ? (How narrow is a laser
beam without one)

Yes. Firstly, all normal radio transmitters are coherent, and they
need large antennas to create directional beams; ordinary daylight or
incandescent light is the exception to coherence, not lasers. Lasers also
need large apertures to create directional beams and they need the phase
to be flat across that aperture. For example, a semiconductor laser actually
produces a beam that diverges over many degrees, but the lenses in a typical
laser pointer form a telescope that converges it into a beam whose divergence
essentially depends on the diameter of the lens. A helium-neon laser, because
of its long cavity, will only allow full amplification for signals that
have stayed within its diameter for a long distance, but is still constrained
by that diameter in determining the lower bound on the beam width.

- How do you project a 1mm laser beam onto a 1m telescope (if you need
one).

I doubt that the sort of lasers that we are talking about would have such
a small aperture, but you would simply add a diverging lens (that's what
the eyepiece does in a simple optical telescope - it provides the whole
area of of the objective lens to the small pupil of the eye, producing
a nominally parallel beam into the eye from a point-like distant source.

Note that your 1mm beam will diverge at least about 1 in 2,000, so would,
for example, be over a 100 miles across at the moon.

- How do you detect ps pulses ?

We are certainly detecting sub-nanosecond pulses as a matter of routine
in optical fibre telephone systems, but these don't have to achieve
ultimate signal to noise ratios. I don't know what the recovery time
and rise time for a typical photmultiplier would be.

- What would happen if an airplane flies over a laser beacon when it's
on, or a bird,

The plane wouldn't be allowed to, the bird might be killed or blinded;
it depends on the power density of the beam, which will be lower for a
highly directional beam. Plane's are not allowed over Arecibo when it
is transmitting and it would be unhealthy for a bird to fly too close
to the feeds, although 30 watts per square metre in the main beam is
probably less than the exposure from a mobile phone.

- Did we do a moon-bounce experiment with laser pulses yet ? Or a
mars-bounce ?

Only with a retro-reflector installed by the Apollo astronauts.

David,

Thank you for a great overview, and your patience to answer all my
newbe-tech-questions on Optical(and microwave) SETI with such
meticulous detail.

Last week, Optical for interstellar communication (including beacons)
seemed the best thing since sliced bread to me : transmitter needs to blast over
the light-output of nearby sun, but with ultra-short pulses of very high
power this is easily within even our current technology. Even disregarding
the bandwidth. And optical is so much easier to beam than these fuzzy
microwave 'photons'.
But after reading some more about it, I realized that the quantum limit is
very real, and very costly. For the receiver, it's not enough to 'see' the
transmitter's star, but each pulse needs to have significant energy to be
detected. Not just power. So receiver needs to be quite large in optical.
So the only thing in optical advantage is it's low-cost transmitter beaming.
But if ET does not know where/if our sun has a ETI planet around it, there
are limitations to the transmitter beam-narrowness. So microwave might
be a lot cheaper than optical to reach nearby stars.
Far-away stars I'm not so sure yet, because microwave transmitter antenna
gets rediculously large.

I want to do some more calculations, and read the SETI 2020
book (I just ordered it on-line), but ultimately, I'd like to make a
solid overview of the parameters involved for interstellar communication,
focusing on minimization of the cost/bit transferred. I'll post my findings
if it turns out that I have something new to report about the
probability and character of finding beacons from other stars.

That brings me to the last unanswered question : How (cost) effective is a
pulsed beacon (optical (ns)or microwave(sec)) versus a continuous beacon. If the
purpose of the beacon is to announce it's presence (very low bit-rate required),
then a pulsed signal seems a lot less costly than a continuous signal.
But if the pulses are too far apart, the receiver will discard the (once in a lifetime)
pulse as a result of noise. Now, I'm sure I learned about S/N ratios for
pulsed signals in Information Theory back in college, but that has been
a while. Does the S/N ratio (power) during the pulse have to go up if you start
separating pulses further and further (change of duty-cycle) ?

Thanks

Rob


"David Woolley" wrote in message ...
In article ,
Rob Dekker wrote:

For example, we have a very nice formula for microwave, which relates
distance, feed-power, antenna size, background noise (system temp),
bandwidth and observation time (signal integration) and such for
recognizable

The only real difference in the formula is that the noise is predominantly
quantum shot noise, rather than thermal noise. (You can also get closer
to 100% illumination efficiency, but the tranmit and receive aperture
parts of the formulae are the same.)

signal reproduction (S/N). I still miss 'pulse-width and pulse-separation'
in
that formula for pulsed signals, but other than that it seems pretty
complete.

Note that pulse width and bandwidth are interrelated, and for pulses less
than about 10 seconds in duration one should be considered almost completely
determined by the other.

There is also a shift between coherent and non-coherent pulse detection when
the pulses are not much more frequent than about 10 per second.

The threshold signal to noise ratio would need increasing, because of the
need to search over many different periods and pulse rates (S@H already has
a high thresold for narrowband signals, because of searching over many chirp
rates).

I'm curious to know if we have (or can create) a similar formula for
optical,
which sets some theoretical limitations of optical communication.

www.coseti.org, as already mentioned, although double check formulaes; I
have reported errors in the past that make it look more favourable.

- How much info can ET put in a pulsed signal ?

Information capacity is not a big issue for SETI. The intention of SETI
is to detect existence. There may be a secondary signal that has high
information content, but the pilot signal has to be easy to detect, rather
than information rich. One proposal is to use a microwave pilot and
an optical main signal.

The most efficient use of power is, probably, to send the transmission
schedule and technical parameters as a second level pilot signal and
cycle through the destinations with the main information signal.

certain things vastly different in optical. We need at least pulse info into
the formula,

Not necessarily. One optical SETI option is to do narrowband detection
using optical superhets. I think the pulse detection case was based on
producing systems that are a relatively cheap add on to high end amateur
scopes.

and background noise is also different : for pulsed laser, the transmitter
defines
most of the S/N ratio, not the receiver, because it needs to blast over the
light output of the sun it is close to.

Not if it uses dark lines in the star's spectrum, even with CW beacons. The
main noise is quantum shot noise.

Anyone knows a good overview of this ?

- Does the transmitter really need a telescope ? (How narrow is a laser
beam without one)

Yes. Firstly, all normal radio transmitters are coherent, and they
need large antennas to create directional beams; ordinary daylight or
incandescent light is the exception to coherence, not lasers. Lasers also
need large apertures to create directional beams and they need the phase
to be flat across that aperture. For example, a semiconductor laser actually
produces a beam that diverges over many degrees, but the lenses in a typical
laser pointer form a telescope that converges it into a beam whose divergence
essentially depends on the diameter of the lens. A helium-neon laser, because
of its long cavity, will only allow full amplification for signals that
have stayed within its diameter for a long distance, but is still constrained
by that diameter in determining the lower bound on the beam width.

- How do you project a 1mm laser beam onto a 1m telescope (if you need
one).

I doubt that the sort of lasers that we are talking about would have such
a small aperture, but you would simply add a diverging lens (that's what
the eyepiece does in a simple optical telescope - it provides the whole
area of of the objective lens to the small pupil of the eye, producing
a nominally parallel beam into the eye from a point-like distant source.

Note that your 1mm beam will diverge at least about 1 in 2,000, so would,
for example, be over a 100 miles across at the moon.

- How do you detect ps pulses ?

We are certainly detecting sub-nanosecond pulses as a matter of routine
in optical fibre telephone systems, but these don't have to achieve
ultimate signal to noise ratios. I don't know what the recovery time
and rise time for a typical photmultiplier would be.

- What would happen if an airplane flies over a laser beacon when it's
on, or a bird,

The plane wouldn't be allowed to, the bird might be killed or blinded;
it depends on the power density of the beam, which will be lower for a
highly directional beam. Plane's are not allowed over Arecibo when it
is transmitting and it would be unhealthy for a bird to fly too close
to the feeds, although 30 watts per square metre in the main beam is
probably less than the exposure from a mobile phone.

- Did we do a moon-bounce experiment with laser pulses yet ? Or a
mars-bounce ?

Only with a retro-reflector installed by the Apollo astronauts.

David,

Thank you for a great overview, and your patience to answer all my
newbe-tech-questions on Optical(and microwave) SETI with such
meticulous detail.

Last week, Optical for interstellar communication (including beacons)
seemed the best thing since sliced bread to me : transmitter needs to blast over
the light-output of nearby sun, but with ultra-short pulses of very high
power this is easily within even our current technology. Even disregarding
the bandwidth. And optical is so much easier to beam than these fuzzy
microwave 'photons'.
But after reading some more about it, I realized that the quantum limit is
very real, and very costly. For the receiver, it's not enough to 'see' the
transmitter's star, but each pulse needs to have significant energy to be
detected. Not just power. So receiver needs to be quite large in optical.
So the only thing in optical advantage is it's low-cost transmitter beaming.
But if ET does not know where/if our sun has a ETI planet around it, there
are limitations to the transmitter beam-narrowness. So microwave might
be a lot cheaper than optical to reach nearby stars.
Far-away stars I'm not so sure yet, because microwave transmitter antenna
gets rediculously large.

I want to do some more calculations, and read the SETI 2020
book (I just ordered it on-line), but ultimately, I'd like to make a
solid overview of the parameters involved for interstellar communication,
focusing on minimization of the cost/bit transferred. I'll post my findings
if it turns out that I have something new to report about the
probability and character of finding beacons from other stars.

That brings me to the last unanswered question : How (cost) effective is a
pulsed beacon (optical (ns)or microwave(sec)) versus a continuous beacon. If the
purpose of the beacon is to announce it's presence (very low bit-rate required),
then a pulsed signal seems a lot less costly than a continuous signal.
But if the pulses are too far apart, the receiver will discard the (once in a lifetime)
pulse as a result of noise. Now, I'm sure I learned about S/N ratios for
pulsed signals in Information Theory back in college, but that has been
a while. Does the S/N ratio (power) during the pulse have to go up if you start
separating pulses further and further (change of duty-cycle) ?

Thanks

Rob


"David Woolley" wrote in message ...
In article ,
Rob Dekker wrote:

For example, we have a very nice formula for microwave, which relates
distance, feed-power, antenna size, background noise (system temp),
bandwidth and observation time (signal integration) and such for
recognizable

The only real difference in the formula is that the noise is predominantly
quantum shot noise, rather than thermal noise. (You can also get closer
to 100% illumination efficiency, but the tranmit and receive aperture
parts of the formulae are the same.)

signal reproduction (S/N). I still miss 'pulse-width and pulse-separation'
in
that formula for pulsed signals, but other than that it seems pretty
complete.

Note that pulse width and bandwidth are interrelated, and for pulses less
than about 10 seconds in duration one should be considered almost completely
determined by the other.

There is also a shift between coherent and non-coherent pulse detection when
the pulses are not much more frequent than about 10 per second.

The threshold signal to noise ratio would need increasing, because of the
need to search over many different periods and pulse rates (S@H already has
a high thresold for narrowband signals, because of searching over many chirp
rates).

I'm curious to know if we have (or can create) a similar formula for
optical,
which sets some theoretical limitations of optical communication.

www.coseti.org, as already mentioned, although double check formulaes; I
have reported errors in the past that make it look more favourable.

- How much info can ET put in a pulsed signal ?

Information capacity is not a big issue for SETI. The intention of SETI
is to detect existence. There may be a secondary signal that has high
information content, but the pilot signal has to be easy to detect, rather
than information rich. One proposal is to use a microwave pilot and
an optical main signal.

The most efficient use of power is, probably, to send the transmission
schedule and technical parameters as a second level pilot signal and
cycle through the destinations with the main information signal.

certain things vastly different in optical. We need at least pulse info into
the formula,

Not necessarily. One optical SETI option is to do narrowband detection
using optical superhets. I think the pulse detection case was based on
producing systems that are a relatively cheap add on to high end amateur
scopes.

and background noise is also different : for pulsed laser, the transmitter
defines
most of the S/N ratio, not the receiver, because it needs to blast over the
light output of the sun it is close to.

Not if it uses dark lines in the star's spectrum, even with CW beacons. The
main noise is quantum shot noise.

Anyone knows a good overview of this ?

- Does the transmitter really need a telescope ? (How narrow is a laser
beam without one)

Yes. Firstly, all normal radio transmitters are coherent, and they
need large antennas to create directional beams; ordinary daylight or
incandescent light is the exception to coherence, not lasers. Lasers also
need large apertures to create directional beams and they need the phase
to be flat across that aperture. For example, a semiconductor laser actually
produces a beam that diverges over many degrees, but the lenses in a typical
laser pointer form a telescope that converges it into a beam whose divergence
essentially depends on the diameter of the lens. A helium-neon laser, because
of its long cavity, will only allow full amplification for signals that
have stayed within its diameter for a long distance, but is still constrained
by that diameter in determining the lower bound on the beam width.

- How do you project a 1mm laser beam onto a 1m telescope (if you need
one).

I doubt that the sort of lasers that we are talking about would have such
a small aperture, but you would simply add a diverging lens (that's what
the eyepiece does in a simple optical telescope - it provides the whole
area of of the objective lens to the small pupil of the eye, producing
a nominally parallel beam into the eye from a point-like distant source.

Note that your 1mm beam will diverge at least about 1 in 2,000, so would,
for example, be over a 100 miles across at the moon.

- How do you detect ps pulses ?

We are certainly detecting sub-nanosecond pulses as a matter of routine
in optical fibre telephone systems, but these don't have to achieve
ultimate signal to noise ratios. I don't know what the recovery time
and rise time for a typical photmultiplier would be.

- What would happen if an airplane flies over a laser beacon when it's
on, or a bird,

The plane wouldn't be allowed to, the bird might be killed or blinded;
it depends on the power density of the beam, which will be lower for a
highly directional beam. Plane's are not allowed over Arecibo when it
is transmitting and it would be unhealthy for a bird to fly too close
to the feeds, although 30 watts per square metre in the main beam is
probably less than the exposure from a mobile phone.

- Did we do a moon-bounce experiment with laser pulses yet ? Or a
mars-bounce ?

Only with a retro-reflector installed by the Apollo astronauts.