archive: Re: SETI Re: [ASTRO] Encounter 2001 "Cosmic Call"

Re: SETI Re: [ASTRO] Encounter 2001 "Cosmic Call"

David Woolley ( )
Tue, 29 Dec 1998 00:45:53 +0000 (GMT)

[ I'm not subscribed to the astro list, although Philip may repost this
reply there. ]

Whilst I think Encounter 2001 is cynical commercialism, and, if detected,
likely to end up as a false negative because of insufficient duration to
allow proper confirmation, I also think that leakage radiation is much less
likely to be detected than a deliberate transmission.

> >
> >To me this entire concept seems silly. My voice has been transmitted to
> >space literally thousands of times when I've used satellite links for phone
> >calls (rarely now since most long distance is via fiber optic), news

There are a number of considerations about satellite uplinks that reduce their
value as leaked signals:

- uplink power will be just sufficient for adequate signal to noise ratio,
both from an economics point of view and because excess power will just
interfere with other uplinks and overload satellite transponders;

- pure frequency components are undesirable, because they overload the
transponder and don't carry information.

The diameter of an antenna needed to recover the modulation is going to go
up in direct proportion to the distance, so a realistic antenna is not
going to be able to recover it at interstellar distances. Pure frequency
components can be enhanced by filtering and integration, but are unlikely
to be a large proportion of the total transmitted power, but only just
enough to serve as pilot frequencies, etc.

> >which don't have more conveinent connections. As a rule the transmitters
> >are very powerful and directional. Only a small percentage of any

But only just powerful enough to do the job.

> >satellite transmission is actually intercepted by the satellite, the
> >remainder continues towards interstellar space.

The high power is obtained by using narrow beams, and for Clarke belt
satellites these are being swept at 15 degrees an hour, which greatly
limits the integration time that can be used.

> >hemisphere in higher technology areas, uplink sites and satellites are not
> >located at regular intervals, the Earth's movement around the sun must also
> >be taken in to account, etc.).

The motion round the sun will produce a parallax effect, but even the nearest
star is approximately 1 parsec away, which means the parallax effect will be
less than about 1 second of arc; which can be ignored, and, in any case,
the receiver would contribute a similar effect. There will be some motion
because of the precession of the earth's axis, but that is of the order of
a degree a century.

> >
> >In addition high power microwave terrestrial transmissions (also less
> >common now due to fiber optics) also 'leak' most of their signal to space
> >instead of reaching their intended destinations (the principle for many
> >classified military radio listening satellites).

Most of these are not likely to contain easily detectable, stable, narrow
band (~1Hz) components.

> >
> >For that matter every time I use my 5 watt handheld amateur transceiver my
> >signal is transmitted to space, although with much less intensity. (enough

The signal from such a transmitter is not at all stable in frequency,
reducing the available sensitivity. It may well be detectable by a
professional radio astronomy receiver from any planet in the solar system,
but is unlikely to show as anything more than excess background noise, when
combined with many other such transmitters, at interstellar distances.

Even non-CDMA digital mobile phones transmit a signal which is near random
(because of vocoding compresssion and probably scrambling) so will not have
very strong pure frequency components (i.e. they will be much more like the
noise from a modem than the 1800 Hz carrier that the modem uses).

> >
> >(BTW did anybody notice in the film "Contact" that the Vegans took almost a

There are no Vegans in the book, and I think not in the film; Vega is a
remote listening outpost.

> >decade before they decided to reply? Berlin Olympics 1936 + 52 years round
> >trip light time makes 1988. The movie took place in 1997 - nine years
> >later.)

That's consistent with when the book was written; the script-writers moved
it to the current day, but probably thought the original reflected event
to have sufficient dramatic effect to retain. Incidentally, I've now
done the power budget calculations, and the antenna required to recover
the modulation would have been gigantic (see below).

> >
> >
> >There are practical limits for the maximum distance a signal can be
> >received before its so attenuated it's below the signal-to-noise floor.
> >Here's a formula for calculating the range for detecting a signal. Radio
> >astronomers, Electrical Engineers, and some ham radio operators should
> >recognize the formula and values. (I believe I got this from the SETI
> >League's web page)
> >
> >
> >Range at which a signal can be detected
> >
> >R=8x10-6*(Pe*A/T)1/2*(t/B)1/4

We have to be careful with this formula. Give or take any missing
factors in the constant, this is based on a probability that the given
measurement was not a random fluctuation that would be obtained by using
the minimum possible averaging time for the bandwidth with a signal power
equal to the average noise power in the bandwidth.

For a single measurement, this is quite a high probability of a false
positive. For multiple measurements, t should really be the total
observation time to date, not that for the single measurement, if one
is to have high confidence of a hit, which means a single measurement at
this level gives virtually no confidence of a detection. I believe the
threshold implied by this formula is a lot smaller than that used for the
initial screening in professional SETI searches and is much much smaller
than what would be needed to declare a detection (although that confidence
might well be obtained by averaging multiple measurements made over the
same averaging time once the system went into a confirmation mode.

> > 4. T is the excess receiver noise temperature in kelvin

For SETI, T includes sky noise.

> >zero. Small yes. Zero no. I've heard that Arecibo could "communicate"
> >with any similar size radio telescope in the Milky Way, although I don't
> >have any reference or evidence to back up that statement.

Lots of assumptions go into such calculations, although Arecibo has one
special constraint, a limited steerability. However, the ability to
communicate (no quotes) depends on the bit rate; for a sufficiently low
bit rate, almost any range is possible, and, if the noise were purely
thermal in nature, that communication could be error free (there will be
some quantum noise and impulsive noise in reality).

Lets take the formula and assume 20dB SNR (to allow the signal to
be demodulated), 10 degrees noise temperature, 100kW EIRP and 4MHz
bandwidth for the Munich signal (SNR as received). t is 1/B as we are
not integrating line to line or frame to frame.


22 = 8E-6 * (1E5 * A / 10) ^ (1/2) * (1 / 4E6) ^ (1/2)

for an SNR of 0 dB.

484 = 8E-6 * 1E4 * A / 100 * (1 / 4E6)

for SNR of 20 dB
484 = A * 2 E-10
A = 242 E10 metres

effective radius of circular antenna = 0.87 E 6 metres. Allow for under
illumination etc., and we get 1,000 km radius, or 2,000 km diameter;
that's certainly much bigger than the United Kingdom. Assuming a 50MHz
carrier, that's a beamwidth of the order of an arc second, although the
range means that the earth's orbit would comfortably fit within the beam
(such an antenna at 1,420 MHz could completely miss the earth if pointed
at the sun).

I don't think these are necessarily incompatible with what was described
in the book.

(Note that actual values for EIRP, carrier frequency and bandwidth should
be available, but I've made guesses. Going from 22 to 26 light years will
increase the size in proportion.)