I'm more than sure this has been brought up on numerous occasions, but surely searching for an intelligent signal in the 'clean' bands is practically a waste of time?
Look at it this way, if there was a civilisation such as ours monitoring our solar system and they used the same technique as us by monitoring the same 'clean' frequencies, they would swiftly come to the conclusion that there was no activity and move on. I believe we have occasional transmitted signals before in these bands, but if in say 50 years when it arrives at the target (assuming it is 50 light years away) they are not monitoring us then they won't hear it.
It makes you wonder, if other civilisations employ similar methods to us, when we believe we have heard a signal and gone back some time later for confirmation and not received it, it may not be that the original is invalid but that they civilisation in question has simply gone back to monitoring.
In a room where everyone is listening, there isn't going to be any conversation...
I share your skeptism about firing off short bursts of messages in the so called clean bands. The probability of anyone listening in our direction and exactly the time the message reaches them is very small (but not zero). My understanding of the choice of using the Water hole as the primary Seti band was that it is a spectral region that Radio Astronomers most frequently use, and thus the transmitter increases this small probability of an accidental detection by an alien astronomer using this band.
I believe that our chances for detecting an alien transmission is greatly improved if some generous race has set up a beacon that is continuously transmitting a signal. However there are good reasons, and some not so good reasons for us not to set up such a beacon in the Sol system.
1) Stephen Hawkins concern that such a beacon could attracts some interstellar bad guys to our system
2) The clean bands now become dirty bands, as we start to transmit many watts of power out to the stars. Any slightest non linearity in the system will have the same effect as the street lights have had on optical astronomy, fogging our radio telescopes with radio pollution.
I, like you, appreciate Frank Drakes nightmare scenario of a Universe full of people just listening. However despite my comments in a previous submission about having doubts regarding picking up ET's leakage from their home world, as they are likely to go Radio silent once they discover the effectiveness of Fibre Optic communication links; my optimism has increased when considering not picking up their domestic RF traffic, but from picking up the stray RF from their interstellar space probes orbiting 'interesting' astronomical objects such as Pulsars etc. (See my last posting to this section of the Forum). Although if they have solved the problem of completely quantum entangling the data acquisition system on the space probe with one on the home world, all bets will be off. Instantaneous data transfer - what a thought.
I'm a newcomer here, but have been around SETI for a very long time (Arecibo Observatory 1974-2000, SETI Institute and ATA, 2000 - 2006).
Now that I'm retired I'll have some time to think about these things.
I liked your idea about the pulsar vicinity housing a beacon, untraceable to its home planet. Ideas like this are one of the reasons the targeted searches have always included some pulsars on the target list. It's a pretty grim engineering and construction environment, given the extremely high energy particles and high magnetic fields, but on the other hand there is lots of energy available in the rotating magnetic fields to power a beacon.
It's fun to think of how you might modulate such a signal. One of my favorites is to orbit the pulsar at just the right distance so that the beacon's pulse rate (orbital period) is some factor like e^pi longer than the pulsar's pulse rate. That would be a dead giveaway that it is a signal of intentional origin.
As to monitoring in the 'clean' bands, I'm just on my last day here at the International Telecommunications Union trying to keep those bands clean. Continuing to participate in Working Party 7D (Radio Astronomy) is one of the things I've kept up in retirement. From my perspective, any advanced civilization willing to go to the trouble of monitoring us (and presumably a large sample of other habitable zone planets) is going to have the wherewithal to monitor -all- frequencies. Their FCC and ITU are going to make different decisions about what bands to keep quiet (if any), but the water hole is indeed a special place. At each end the ITU has designated
'do not transmit', but this is only 1400-1427 MHz (H, neutral hydrogen) and 1660 - 1670 MHz (OH, the other component of water). In between you have Iridium, GPS and almost all the L-band satellites. It's hardly 'sticking to the clean bands'. The amazing thing is that it seems to be feasible to work in the water hole, using sophisticated pattern recognition to isolate terrestrial signals. About 80% is accessible at least part of the time, in spite of the millions of dollars per MHz invested their by the communications industry.
The protected band is only 1400-1427 MHz? I had no idea it was this small. Any nearby wideband modulated non-linearity will spill right in. This could explain some things in the datasets.
Your concept of putting the beacon at the e^pi orbit rather captured my imagination, so I done some back of the envelop calculations, based on several slightly invalid assumptions
1) I could use Kepler's laws regarding orbital periods - I know that one should use General Relativity in such a high gravitational field, alas my envelop wasn't big enough.
2) I assumed that the stellar remnant was approximately the mass of the sun, again this is an under estimate, I vaguely remember that the limits for a supernova core can go one of 3 ways, less than Chandresakars(?) limit it becomes a white dwarf, very much higher than this limit it becomes a black hole; anywhere between these limits it becomes a neutron star. Unfortunately I have forgotten where these limits actually are.
3) I assumed that the Pulsar period was about 0.7 seconds.
With these assumptions your Orbital period comes out at about 16 seconds, with the probe at about 6000 miles above the Pulsar centre, so it should be clear of the surface. It will have an orbital velocity about 0.01 x the speed of light, so for a relativistic probe, this should be well within its capability. When I have another envelop it will be instructive to some calculations on the Doppler effect on the Beacons signal.
My own thought would be to put the probe on a high eccentricity orbit, so you get the chance to investigate the smashed planetary system, if it has one, as well as the cores vicinity. I guess it would be safer to have the orbital plane somewhat out of the ecliptic, to keep it clear from planetary debris, which I guess wouldn't do the probe much good if you had a collision. But on the other hand I guess you wouldn't want a Polar orbit, to keep you clear from the jets generating the pulses.
Anyway just a few thoughts
Although the concept of putting a probe into the e^pi orbit would certainly provide a nice clear indication that the probe was placed in orbit by an intelligent entity. I am not sure a better signal frequency would be one that was a direct multiple of the Pulsar frequency. My thinking behind this is that knowing the exact repeat frequency of the Pulsar one can undertake noise reduction using what I know as Synchronous Integration whereby the signal from the Pulsar is broken into durations that are exact integer number of Pulsar periods long, and then these are summed, the signal to noise ratio improves as the square root of the number of data frames that are included in the integration. There is no reason why the observation of the Pulsar could not be continued for extremely long period of time (e.g. 10's of years) thus removing virtually all of the noise, and leaving the signal from the probe, even if its transmission power were minimal, visible to us. Having an array telescope we need not keep the complete system tied up in tracking the Pulsar at the expense of any other research, we can use Aperture synthesis to generate a secondary beam targeted on our Pulsar whilst leaving the main beam doing the current work. If the observation train gets broken (e.g. the air conditioning stops working;-) all is not lost, for all we need to do is once observation is restarted we can use something like correlation to calculate the relative phase of the new set of data from the old set using the Pulsar signal itself. Once any shift is compensated for, the new set of data can be continued to be added into the on going accumulation.
The main problem I can see with this technique is how to indicate the signal from the probe is just that, and not some fine detail in Pulsars signature. My own guess is that in addition to the main probe pulse one could add a secondary pulse or two, and modulate the position of these secondary pulses wrt to the main probe pulse (at a relatively slow rate, so that the long time integration required to pull the signal out would be able to resolve both sets of pulses, without losing the time drifting one back into the noise).
If there is anyone left out there, apart from me, still reading this stuff I would be glad of some feedback
Those are interesting ideas. Assuming a wide-band pulse signal that has the same directionality as the pulsar's beam, here are some questions:
After reading your posts I began to see the magnetic fields of your probe-pulsar system as a giant alternator in the sky. Just like in an automobile, this alternator would want to generate AC current. We could have enormous rectifying diodes and a massive capacitor to store the energy for a wide-band pulse. Or we could use this AC source directly as a low frequency oscillator to transmit a narrow band signal that could be modulated. I'm just thinking that it's easier and a lot more efficient if we can avoid having to convert AC to DC at such massive energies. This would defeat your synchronous integration idea but it would definitely look artificial. Especially if we placed several of these "probes" at different orbiting distances or spinning rotational rates which would produce different AC transmitting frequencies.
With regards to the Signal to Noise improvement I was expecting from using Synchronous Integration; I did the following calculation based on a Hypothetical Pulsar round which an Alien Species had orbited a probe. I had made the following assumptions about the Pulsar:-
We need to ensure that we have sufficient arithmetic headroom, that the accumulation will neither saturate, or else loose the incoming signal to truncation, so this number of observations will require an arithmetic headroom of approximately 30 bits, so with an 8 bit ADC this would require at least 38 bits precision, so a modern 64 bit machine should be able to cope with this requirement.
Now I made a really unjustifiable assumption that I could model the noise as simple additive Gaussian interference (No I really don't really believe the environment from the Pulsar is going to be quite that simple - but I only had a small envelope) So using that model I can calculate the improvement in SNR by finding the squareroot of the number of observations, generating an answer that we could extract the signal if it were approximately 2.9x10^4 below noise level.
I would be interested in your comments, I was slightly surprised that over such a long protracted observation period the answer wasn't greater, although I have to admit that I could have goofed in my arithmetic.
I rather like you AC signal concept. Not sure how I would go about modelling it though.
by finding the squareroot of the number of observations, generating an answer that we could extract the signal if it were approximately 2.9x10^4 below noise level.
I would be interested in your comments, I was slightly surprised that over such a long protracted observation period the answer wasn't greater,
The SNR improvement goes as sqrt(t * BW), where t is observation time in seconds and BW is bandwidth in Hertz. The pulsar period is a term, but in its ratio to pulse width (see equation in book image below). One way to think of this is that a slowly rotating pulsar pulse will be observed less often but there will be more time samples of the pulse that can be averaged (noise reduced) to reach the equivalent smaller time resolution (less samples) for the more frequently observed pulse of a faster rotating pulsar (for two pulsars of roughly the same pulse width to period ratio).
Search the Internet for: pulsar sensitivity radiometer equation
I agree that what Dave suggested should be coherent integration with the SNR gains Rob mentioned but has anyone actually done this with a pulsar over a long duration? The theory is great but I see all sorts of technical challenges.
Some reasons I ask are:
* Pulsars spin down over time, their period changes but this could be accounted for. How accurately and with what sort of errors?
* Pulsars glitch, what sort of noise does this introduce?
* I suspect a pulsar is an extremely harsh and noisy environment. What is the background noise floor of a pulsar? What are its characteristics? It would likely be a function of the spin rate which means we would coherently integrate it.