Literature on remote sensing

[Rune Allnor]

Hi all.

I am involved in a project where we want to measure a
physical quantity by means of acoustics. We have tried
ultrasound pulse signals with a certain amount of
success, but it seems we have pushed the system
to the edge of the envelope. So we need to explore other
options to reach the goal of the project.

My background is from seismics and underwater
acoustics where all attention is focused on pulsed
signals. I need to read up on other ways of doing
things. I presume "other ways" mainly means "radar",
but "tomography" may also be relevant.

Basically, I need tips on literature. The ideal book(s)
would be general in scope but on an advanced level.

I have heard of the Skolnik book, but I can't find it.
Well, I can, but the book is a bit dated (it was
published in the mid 60s?), and I would like to hear
if there are newer texts on general radar processing
available before spending time and $$ on finding the
Skolnik text.

Rune

[Steve Underwood]

Post by Rune Allnor
Hi all.
I am involved in a project where we want to measure a
physical quantity by means of acoustics. We have tried
ultrasound pulse signals with a certain amount of
success, but it seems we have pushed the system
to the edge of the envelope. So we need to explore other
options to reach the goal of the project.
My background is from seismics and underwater
acoustics where all attention is focused on pulsed
signals. I need to read up on other ways of doing
things. I presume "other ways" mainly means "radar",
but "tomography" may also be relevant.
Basically, I need tips on literature. The ideal book(s)
would be general in scope but on an advanced level.
I have heard of the Skolnik book, but I can't find it.
Well, I can, but the book is a bit dated (it was
published in the mid 60s?), and I would like to hear
if there are newer texts on general radar processing
available before spending time and $$ on finding the
Skolnik text.

The second edition of the Skolnik book was published around 1980, and
was substantially updated. However, it is still heavily biased towards
radars which emit pulses. I seem to remember it covers FM techniques a
bit, and completely fails to address things like the transmission of
wideband noise as a stimulus.

I haven't worked in the radar area since the mid 80s, so I don't know
much about more modern texts. However, any in depth book on radar is
biased very much towards stealth, since detecting targets is the easy
part, and LPI is where it starts to get interesting. :-) If you are
targeting a civilian application they probably don't have a lot to teach
you.

You didn't actually say what limits you have reached. Is it range,
resolution, sensitivity, etc.? That etc. can be the one that really gets
you. :-)

Regards,
Steve

[Rune Allnor]

Post by Steve Underwood

Post by Rune Allnor
Hi all.
I am involved in a project where we want to measure a
physical quantity by means of acoustics. We have tried
ultrasound pulse signals with a certain amount of
success, but it seems we have pushed the system
to the edge of the envelope. So we need to explore other
options to reach the goal of the project.
My background is from seismics and underwater
acoustics where all attention is focused on pulsed
signals. I need to read up on other ways of doing
things. I presume "other ways" mainly means "radar",
but "tomography" may also be relevant.
Basically, I need tips on literature. The ideal book(s)
would be general in scope but on an advanced level.
I have heard of the Skolnik book, but I can't find it.
Well, I can, but the book is a bit dated (it was
published in the mid 60s?), and I would like to hear
if there are newer texts on general radar processing
available before spending time and $$ on finding the
Skolnik text.

The second edition of the Skolnik book was published around 1980, and
was substantially updated. However, it is still heavily biased towards
radars which emit pulses. I seem to remember it covers FM techniques a
bit, and completely fails to address things like the transmission of
wideband noise as a stimulus.
I haven't worked in the radar area since the mid 80s, so I don't know
much about more modern texts. However, any in depth book on radar is
biased very much towards stealth, since detecting targets is the easy
part, and LPI is where it starts to get interesting. :-) If you are
targeting a civilian application they probably don't have a lot to teach
you.
You didn't actually say what limits you have reached. Is it range,
resolution, sensitivity, etc.? That etc. can be the one that really gets
you. :-)

All of the above...

We are working in a noisy area (acoustically, electrically) and in
a high-loss, modest-coherence material and try to detect relatively
weak signals with as high precision and at as long range we can.
We can, of course, spend tens of thousands of $$ on high-power
amplifiers and transducers, but I can't really see any benefit
of that. So I want to try to increase signal energy by increasing the
time-bandwidth product of the signal rather than the emitted power
of the signal.

Something that covers frequency-sweep radar ought to do nicely
for this application. Anything more than that is at least interesting
reading, maybe an idea for the next project.

Rune

[Leif Harcke]

Post by Rune Allnor
So I want to try to increase signal energy by increasing the
time-bandwidth product of the signal rather than the emitted power
of the signal.

Most radar books contain a chapter on "pulse compression," which is the
radar jargon for increasing the pulse energy while keeping the peak pulse
power constant, and without decreasing the achievable time/range
resolution.

The classic text is the 1967 volume "Radar Signals" by Cook and Bernfeld.
Levanon and Mozeson have written a new book with the same old title,
"Radar Signals," which Wiley released in 2004.

-Leif

[Rune Allnor]

Post by Leif Harcke

Post by Rune Allnor
So I want to try to increase signal energy by increasing the
time-bandwidth product of the signal rather than the emitted power
of the signal.

Most radar books contain a chapter on "pulse compression," which is the
radar jargon for increasing the pulse energy while keeping the peak pulse
power constant, and without decreasing the achievable time/range
resolution.

The idea has occured to me. The problem is that the physical scales
and pulse lengths are on roughly the same scale. The kit we have
available
can send an FM pulse of duration T. The propagation times are in the
range 1.5T - 2T. Not ideal, but the next thing to check.

Post by Leif Harcke
The classic text is the 1967 volume "Radar Signals" by Cook and Bernfeld.
Levanon and Mozeson have written a new book with the same old title,
"Radar Signals," which Wiley released in 2004.

Thanks. I'll see if I can find it.

Rune

[Steve Underwood]

Post by Rune Allnor

Post by Steve Underwood

Post by Rune Allnor
Hi all.
I am involved in a project where we want to measure a
physical quantity by means of acoustics. We have tried
ultrasound pulse signals with a certain amount of
success, but it seems we have pushed the system
to the edge of the envelope. So we need to explore other
options to reach the goal of the project.
My background is from seismics and underwater
acoustics where all attention is focused on pulsed
signals. I need to read up on other ways of doing
things. I presume "other ways" mainly means "radar",
but "tomography" may also be relevant.
Basically, I need tips on literature. The ideal book(s)
would be general in scope but on an advanced level.
I have heard of the Skolnik book, but I can't find it.
Well, I can, but the book is a bit dated (it was
published in the mid 60s?), and I would like to hear
if there are newer texts on general radar processing
available before spending time and $$ on finding the
Skolnik text.

The second edition of the Skolnik book was published around 1980, and
was substantially updated. However, it is still heavily biased towards
radars which emit pulses. I seem to remember it covers FM techniques a
bit, and completely fails to address things like the transmission of
wideband noise as a stimulus.
I haven't worked in the radar area since the mid 80s, so I don't know
much about more modern texts. However, any in depth book on radar is
biased very much towards stealth, since detecting targets is the easy
part, and LPI is where it starts to get interesting. :-) If you are
targeting a civilian application they probably don't have a lot to teach
you.
You didn't actually say what limits you have reached. Is it range,
resolution, sensitivity, etc.? That etc. can be the one that really gets
you. :-)

All of the above...
We are working in a noisy area (acoustically, electrically) and in
a high-loss, modest-coherence material and try to detect relatively
weak signals with as high precision and at as long range we can.
We can, of course, spend tens of thousands of $$ on high-power
amplifiers and transducers, but I can't really see any benefit
of that. So I want to try to increase signal energy by increasing the
time-bandwidth product of the signal rather than the emitted power
of the signal.
Something that covers frequency-sweep radar ought to do nicely
for this application. Anything more than that is at least interesting
reading, maybe an idea for the next project.
Rune

Chirping certainly cuts the peak power, and has been almost universal in
pulsed radars since the middle of the second World War. They got their
output powers up to the point where flashover in the waveguides forced a
rethink of brute force narrow pulses. It sounds like your issues are
similar. Surprisingly it took many years before a proper mathematical
analysis of the topic was available. In my limited experience of sonar,
chirping also seems pretty normal there.

Some form of continuous transmission really seems the way to go there
days, though. Either FM schemes, or the transmission of wideband noise.
FM has been used for a long time in radar, with a mix of + and -
qualities over chirping (e.g. the Foxhunter radar used on many Tornado
fighters is a continuous transmission FM system). Noise based radars
were in the research stage last time I worked on radar. The research
systems were offering very interesting results, so I guess that has
worked its way into production systems now the compute requirement is
fairly easy to meet.

Regards,
Steve

[Jerry Avins]

Steve Underwood wrote:

...

Post by Steve Underwood
Some form of continuous transmission really seems the way to go there
days, though. Either FM schemes, or the transmission of wideband noise.
FM has been used for a long time in radar, with a mix of + and -
qualities over chirping (e.g. the Foxhunter radar used on many Tornado
fighters is a continuous transmission FM system). Noise based radars
were in the research stage last time I worked on radar. The research
systems were offering very interesting results, so I guess that has
worked its way into production systems now the compute requirement is
fairly easy to meet.

Please educate me. The T/R tube was one of the finest bits of invention
in early radar, protecting the receiver's sensitive input from the
transmitter's blast of power. How does the receiver cope with continuous
transmission?

Jerry

[Steve Underwood]

Post by Jerry Avins
...

Post by Steve Underwood
Some form of continuous transmission really seems the way to go there
days, though. Either FM schemes, or the transmission of wideband
noise. FM has been used for a long time in radar, with a mix of + and
- qualities over chirping (e.g. the Foxhunter radar used on many
Tornado fighters is a continuous transmission FM system). Noise based
radars were in the research stage last time I worked on radar. The
research systems were offering very interesting results, so I guess
that has worked its way into production systems now the compute
requirement is fairly easy to meet.

Please educate me. The T/R tube was one of the finest bits of invention
in early radar, protecting the receiver's sensitive input from the
transmitter's blast of power. How does the receiver cope with continuous
transmission?
Jerry

Well these days for pulsed systems you can get maybe 60dB isolation with
a high quality circulator at microwave frequencies, and then use PIN
diodes to protect the receiver. Great until someone drops something
metal in front of the antenna. Then, the PIN diodes can turn a little
spectacular. :-)

For continuous transmission things are more interesting. You have less
intense power, and a circulator will still give you 60dB. That means the
transmitter won't blow the receiver apart. Actually making things work
is left as an exercise for the reader.

Steve

[Jerry Avins]

Post by Steve Underwood

Post by Jerry Avins
...

Post by Steve Underwood
Some form of continuous transmission really seems the way to go there
days, though. Either FM schemes, or the transmission of wideband
noise. FM has been used for a long time in radar, with a mix of + and
- qualities over chirping (e.g. the Foxhunter radar used on many
Tornado fighters is a continuous transmission FM system). Noise based
radars were in the research stage last time I worked on radar. The
research systems were offering very interesting results, so I guess
that has worked its way into production systems now the compute
requirement is fairly easy to meet.

Please educate me. The T/R tube was one of the finest bits of
invention in early radar, protecting the receiver's sensitive input
from the transmitter's blast of power. How does the receiver cope with
continuous transmission?
Jerry

Well these days for pulsed systems you can get maybe 60dB isolation with
a high quality circulator at microwave frequencies, and then use PIN
diodes to protect the receiver. Great until someone drops something
metal in front of the antenna. Then, the PIN diodes can turn a little
spectacular. :-)
For continuous transmission things are more interesting. You have less
intense power, and a circulator will still give you 60dB. That means the
transmitter won't blow the receiver apart. Actually making things work
is left as an exercise for the reader.

Blowing the receiver apart isn't the only problem. Even 60 dB down is a
whale of a lot of co-channel interference.

I once used a Gunn diode and a horn to make a traffic radar (to prove to
the local cops that there was a problem they didn't want to know about).
The return signal simply beat with the diode's output, and the beat
signal went to a counter. I suppose Other continuous radars might work
the same way, but I can imagine lits of issues, There are probably even
more issues I can't imagine.

Jerry

[Steve Underwood]

Post by Jerry Avins

Post by Steve Underwood

Post by Jerry Avins
...

Post by Steve Underwood
Some form of continuous transmission really seems the way to go
there days, though. Either FM schemes, or the transmission of
wideband noise. FM has been used for a long time in radar, with a
mix of + and - qualities over chirping (e.g. the Foxhunter radar
used on many Tornado fighters is a continuous transmission FM
system). Noise based radars were in the research stage last time I
worked on radar. The research systems were offering very interesting
results, so I guess that has worked its way into production systems
now the compute requirement is fairly easy to meet.

Please educate me. The T/R tube was one of the finest bits of
invention in early radar, protecting the receiver's sensitive input
from the transmitter's blast of power. How does the receiver cope
with continuous transmission?
Jerry

Well these days for pulsed systems you can get maybe 60dB isolation
with a high quality circulator at microwave frequencies, and then use
PIN diodes to protect the receiver. Great until someone drops
something metal in front of the antenna. Then, the PIN diodes can turn
a little spectacular. :-)
For continuous transmission things are more interesting. You have less
intense power, and a circulator will still give you 60dB. That means
the transmitter won't blow the receiver apart. Actually making things
work is left as an exercise for the reader.

Blowing the receiver apart isn't the only problem. Even 60 dB down is a
whale of a lot of co-channel interference.
I once used a Gunn diode and a horn to make a traffic radar (to prove to
the local cops that there was a problem they didn't want to know about).
The return signal simply beat with the diode's output, and the beat
signal went to a counter. I suppose Other continuous radars might work
the same way, but I can imagine lits of issues, There are probably even
more issues I can't imagine.
Jerry

Of course. The received signal is at about the same frequency as the
transmitted one, so the full power of the transmitter, less the modest
attenuation you can achieve, is interferring with a tiny kTB limited
signal. Its like the CDMA near/far problem on steroids. As I said,
methods to deal with this are left as an exercise for the reader.

Remember: nothing is impossible - there's always kTB :-)

Regards,
Steve

[Jerry Avins]

... As I said, methods to deal with this are left as an exercise for
the reader.
Remember: nothing is impossible - there's always kTB :-)

Well, as Tom Lehrer had Wernher Von Braun say, "That's not my
department."* (And a good thing too!)

Jerry
__________________________________________
* Once the rockets go up, who cares where they come down.
"That's not my department" says Wernher Von Braun.

[Rune Allnor]

Post by Steve Underwood

Post by Rune Allnor

Post by Steve Underwood

Post by Rune Allnor
Hi all.
I am involved in a project where we want to measure a
physical quantity by means of acoustics. We have tried
ultrasound pulse signals with a certain amount of
success, but it seems we have pushed the system
to the edge of the envelope. So we need to explore other
options to reach the goal of the project.
My background is from seismics and underwater
acoustics where all attention is focused on pulsed
signals. I need to read up on other ways of doing
things. I presume "other ways" mainly means "radar",
but "tomography" may also be relevant.
Basically, I need tips on literature. The ideal book(s)
would be general in scope but on an advanced level.
I have heard of the Skolnik book, but I can't find it.
Well, I can, but the book is a bit dated (it was
published in the mid 60s?), and I would like to hear
if there are newer texts on general radar processing
available before spending time and $$ on finding the
Skolnik text.

The second edition of the Skolnik book was published around 1980, and
was substantially updated. However, it is still heavily biased towards
radars which emit pulses. I seem to remember it covers FM techniques a
bit, and completely fails to address things like the transmission of
wideband noise as a stimulus.
I haven't worked in the radar area since the mid 80s, so I don't know
much about more modern texts. However, any in depth book on radar is
biased very much towards stealth, since detecting targets is the easy
part, and LPI is where it starts to get interesting. :-) If you are
targeting a civilian application they probably don't have a lot to teach
you.
You didn't actually say what limits you have reached. Is it range,
resolution, sensitivity, etc.? That etc. can be the one that really gets
you. :-)

All of the above...
We are working in a noisy area (acoustically, electrically) and in
a high-loss, modest-coherence material and try to detect relatively
weak signals with as high precision and at as long range we can.
We can, of course, spend tens of thousands of $$ on high-power
amplifiers and transducers, but I can't really see any benefit
of that. So I want to try to increase signal energy by increasing the
time-bandwidth product of the signal rather than the emitted power
of the signal.
Something that covers frequency-sweep radar ought to do nicely
for this application. Anything more than that is at least interesting
reading, maybe an idea for the next project.
Rune

Chirping certainly cuts the peak power, and has been almost universal in
pulsed radars since the middle of the second World War. They got their
output powers up to the point where flashover in the waveguides forced a
rethink of brute force narrow pulses. It sounds like your issues are
similar. Surprisingly it took many years before a proper mathematical
analysis of the topic was available. In my limited experience of sonar,
chirping also seems pretty normal there.
Some form of continuous transmission really seems the way to go there
days, though. Either FM schemes, or the transmission of wideband noise.
FM has been used for a long time in radar, with a mix of + and -
qualities over chirping (e.g. the Foxhunter radar used on many Tornado
fighters is a continuous transmission FM system).

Maybe that's what I am looking for. I have heard of radars who use a
"frequency sweep" signal to measure the standing wave response of the
(stationary) environment, and then uses some Fourier transform method
to generate the actual images. In our application, the target is
stationary,
for all practical purposes, so this may be a way to go. If the Tx/Rx
crosstalk
problem can be handled in a sensible way, that is...

Post by Steve Underwood
Noise based radars
were in the research stage last time I worked on radar. The research
systems were offering very interesting results, so I guess that has
worked its way into production systems now the compute requirement is
fairly easy to meet.

Rune

[Steve Underwood]

Post by Rune Allnor
[...]
Maybe that's what I am looking for. I have heard of radars who use a
"frequency sweep" signal to measure the standing wave response of the
(stationary) environment, and then uses some Fourier transform method
to generate the actual images. In our application, the target is
stationary,
for all practical purposes, so this may be a way to go. If the Tx/Rx
crosstalk
problem can be handled in a sensible way, that is...

One FM technique (I think the commonest) is to send out a transmission
which sweeps in frequency from a while, then sticks at one frequency for
a while, then repeats the sweep, etc. During the sweep, the frequency
difference between the transmitted and received signal is a measure of
the target range and resolved velocity munged together. To find the
doppler due to the resolved velocity you look at the frequency
difference during the non-sweeping periods. Then you can remove that
component from the frequency difference during the sweeps, and determine
the actual range of the target. That sounds straightforward, but when
you have more than one target within the beam it gets messier.

Regards,
Steve