Cold fusion is impossible and Physics have long demostrated it.
Nothing is impossible.
Both of these statements are silly.
What we _can_ say about P&F "cold fusion" is that we have not been able, despite much trying, to produce convincing evidence of nuclear fusion in apparatus along the lines of the P&F "cold fusion" cells. The anomalous heat production almost certainly comes from something else.
As far as "nothing is impossible" is concerned, while from a philosophical perspective that statement is true, in practice, some physical laws (most of the ones taught in the textbooks that you ridicule, for instance) have been tested thoroughly enough that it's extremely unlikely they'll be violated in their domain of applicability. Step off a cliff, and you'd better have a bungee cord handy, because gravity isn't going to be repealed any time soon. Similarly, we've been studying nuclear effects for pretty much all of the 20th century. We know with confidence how excited nuclei drop to their ground states, and the fact that we haven't seen evidence of this in P&F "cold fusion" cells shows pretty conclusively that whatever's happening isn't fusion.
They didn't fabricate results, their results just became public too quickly
Oh? I seem to recall hearing about a neutron emission energy spectrum plot with a peak that kept wandering around between press conferences, until they finally withdrew it.
I'm going to have to pick up a copy of "Yes, We Have No Neutrons" one of these days so that I can have all of the questionable bits at my fingertips for situations like this.
I can't give information on the articles, but I can answer some of your questions:
1) Is that f not absorbed in space? What? Isn't space filled with H (relative to other elements).
Space is "filled with H" in that H is the least scarce element in space. What space is mostly filled with is a whole lot of nothing - laboratory-grade vacuum doesn't even come close to being as empty as most of space. So, radio signals (and light, for that matter) propagate vast distances without much trouble.
In the plane of the galactic disc, there's enough material that we can't actually see clear through the galaxy, but there are windows in the spectrum where absorption is much less, and that's probably where SETI is looking. Check the SETI page to see if they have details on the actual bands they're listening at.
2) How powerful must the xmiter be? That may tell alot.
This depends on how far away it is, how directional it is, how much noise your detector sees, and how you're trying to interpret the signal. A 1W source at a distance of 1 LY gives about 1e-33 W/m^2, if radiating in all directions. One radio photon has on the order of 1e-25 J (varies considerably with frequency). If you're listening with a dish with 1 m^2 area, you'll see an average one photon every 1e+8 seconds (3 years). That's how long you'd have to wait to determine whether the signal was present or absent in the complete absence of noise, which means you could pick up modulations of 1e-8 Hz at most, even under perfect conditions.
In practice, you have a noise floor which you have to rise above, and are looking for modulations in the MHz or greater range, and are looking at more distant sources, so power is much greater. Magnetic storms on stars are one of the main signal sources, for a sense of scale.
Of course, a walkie-talkie next to the dish, or an earthbound electrical storm anywhere nearby, would be picked up too. This is why many measurements are needed to verify a celestial emission source, and why radio telescopes are in remote areas.
3) what about com. by gravity? Is gravity instantaneous? though subject to r^2.
Gravity travels at the speed of light. This is why gravity waves exist.
5) We did signal back with the same signal followed by "hello" followed by the same signal. Didn't we? why not?
Because these telescopes are receivers, not transmitters, and because any response would take centuries or longer to reach its destination, and because the signal probably had an earth-based source (noise happens - a lot).
6) BTW, what would be the frequency signature of an H bomb reflected from a far away moon?
An h-bomb set off in space mainly gives off a pulse of hard radiation. If a moon is nearby, you might get x-ray backscatter from Bremsstrahlung radiation. If an object with an atmosphere is nearby, ionizing radiation striking the atmosphere may create lightning (as with very-powerful bombs set off on earth), which will give you radio noise that sounds much like any other electrical storm would (though it would differ by only being present briefly).
The flash and radiation pulse of the bomb itself would be brighter, but as it would have fewer photons, it might not be as easily detected far away. Either way, an H-bomb probably wouldn't be visible from another star system (just not enough photons produced). Communication between stars pretty much requires beamed communication, unless truly massive amounts of energy are involved.
As yet however, it is my understanding that every extra solar planet discovered has been by detecting the gravitational wobble rather than direct observation. As we have both said "in the future"(given enough money/time) it should be possible.
I will begin to change my views once we actually *see* (radio or optical) them with our own eyes though:)
Running numbers, it looks like you'd only need a radio telescope array on the order of a few kilometres wide to resolve planets as distinct from their host stars within 100 LY or so. It seems odd that we haven't tried to directly view nearby superjovian planets this way.
My speculation is that it's because we'd need a very large filled-aperture dish to collect enough photons to sift the desired signal out of the background, but I'd be interested if any radio astronomers lurking could tell me the real constraints on trying this (as we know more or less where the target planets are).
I ought to know enough about discrete ergodic theory to know the answer to this, but is there a simple definition of "entropy" in this situation?
The short answer is "no"; it depends on how you decode the signal into a symbol stream, and several other parameters. However, most signals of natural origin have noiselike content no matter how you decode them, giving high entropy no matter how you measure it. An artificial signal that's intended to be easily picked up would be designed to have low entropy when interpreted in a very basic way (though this could be made more complex as an IQ test, per one of my other posts).
Similar problems with "the frequency spectrum" - just trying to decide on a time domain sampling interval [prior to applying a Fourier Transform] is a whale of a problem in and of itself.
The bandwidth and intensity of your signal impose limits on the data rates that may be present (due to photon count, and other environment-induced noise). This would let you pick reasonable parameters for frequency-domain analysis given the strength and bandwidth of the received signal. SETI's already doing something like this, I believe. The only major problem I can see offhand is that a signal may be modulated too _slowly_ to recognize with a small window, but I don't consider that a likely scenario if the signal is intended to be understood (they'd modulate it as quickly as possible while still having reasonable noise resistance).
As an aside, I'd ask you the same questions I asked another poster: Doesn't a Pulsating Radio Source [PULSAR] have a nice, stable, utterly utterly predictable spectrum? Similarly, doesn't 120V/60Hz [cf 240V/50Hz] wall current have a nice, stable, utterly predictable spectrum? Now which is interesting, and which is uninteresting? And does either one of them transmit a "message"? Or are they more akin to noise?
This question has confused astronomers as well - it was conjectured that pulsars were artificial beacons, until a natural explanation of them arose. An artificial signal intended to be picked up would be made unambiguous by encoding information that would not be generated by natural processes under any conditions. Encoding the first handful of prime numbers is a good way of doing this (either as binary, or as a series of delays between pulses, or what-have-you). Similarly, emitting a narrow-band signal at a small multiple or simple fraction of the wavelength of a naturally occurring spectral line would be recognizable as artificial. A repeating pulse is artificial-looking, but not artificial-looking enough to be unambiguous. So, I'd expect a beacon to have a more blatantly artificial pattern.
Most noise sources have very high entropy, and spectra that fall into a few well-defined shapes.
What are these "few well-defined shapes" [again, presumably relative to which "well-defined sampling intervals"]?
Flat spectra for white noise, 1/f spectra for "pink noise" or "flicker noise", and things like Poisson and normal distributions. Sampling interval doesn't really matter, here, so I'm not sure why you're bringing it up - think of this discussion as referring to continuous-time aperiodic transforms. Changing the scaling factor you're using alters the size but not the shape of the frequency spectrum that results.
Look up texts for semiconductor device physics for discussions of a number of examples of these types of noise come from. Celestial noise sources come from very different sources, but they still tend to follow the same types of distribution (as they arise naturally from the action of various simple classes of random processes - see a statistics text for a discussion of this).
The spectra of artificial signals is radically different, at least if produced by simple equipment or intended to be interpreted by simple receivers. You get a pair of spikes, lobes, or top-hats for a signal modulated onto a carrier. If you're deliberately making the signal as artificial-look
A chaotic system is *not* a random one, because, as you point out, given identical starting conditions, you get identical outputs. It's completely deterministic. I've still yet to see a convincing argument that randomness is more than a useful model for a set of deterministic processes we don't yet understand fully.
Look up "hidden variables" interpretations of quantum mechanics for discussions about this. The upshot is that while it's very difficult to prove the case one way or the other (and there are still people trying), you end up with a far cleaner model if quantum processes are treated as truly random. From there, in the absence of evidence of determinism, Occam's Razor kicks in.
I vaguely recall that there were actually experiments that ruled out some of the simpler "hidden variables" interpretations, strengthening the case for true randomness in quantum events, but you'll have to ask someone other than me for the details.
See, that's my problem. I'd like to go with your EM theory of communication, but I can't find any way of telling whether it's biased. So, how can I tell if what I've picked is subject to bias or not?
Please spell out how! That's all I'm asking.
The "can't possibly be biased" element from my original post was the frequency of the light being transmitted, not the use of EM in the first place. Hydrogen emission lines are fundamental constants of nature (at least in our neck of the woods; ObCaveats about very, very distant parts of the universe possibly having different values for these constants). Choosing a blatant derivative of one would be an unambiguous way of saying "look, I'm intelligent!" that is not tainted by human biases (except in the definition of "intelligence").
The argument for EM is that we seem to understand virtually all of the observable universe now, unlike your proposed witchcraft-using primitives, and EM looks like the _only_ choice for long-distance communication. The only possible objection is that we can't be sure that we understand the universe well enough to label alternate communications modes "unlikely".
I consider it likely that we understand the universe well enough to do this, as the relevant bits of physical theory have been around arguably even since Newton's day, and have stood despite very aggressive challenging (new models are best viewed as extending, not overturning, old, as they tend to reduce to the old models in the old models' domains of applicability). Your opinion may vary (and apparently does).
Right. Let's say we're broadcasting to a target audience of aliens at our level of civilization, say, between mid-20th-century and 10,000 years in the future (any more advanced, and they're not going to even bother... more important things to do).
So we're looking for aliens in a 10,000 year window. The universe is 15 billion years old, give or take. It sounds unlikely to me.
There are on the order of a hundred billion stars in our galaxy. With ObCaveats about assuming a uniform distribution across the 15 billion year domain, that still gives us an estimated 70,000 or so civilizations that are open to contact in our metaphorical backyard. We wouldn't get a timely answer from most of them, but my point is that they're still there, even with the numbers you postulate.
IMO, it's worth a shot. It costs us very little to look, and we learn something either way.
Its not a problem of the actual signals degrading.
They get completely overpowered by the huge great big solar radio emitter, so that by the time they reach another starsystem, all thats resolvable is the signal from our sun itself.
This actually turns out not to be the case, for a couple of reasons. First, Earth outshines the sun on several radio bands - the sun's dumping most of its energy as visible light, and while electrical effects in its atmosphere are noisy, they don't cover the entire radio spectrum. Second, we could launch solar-orbit radio telescope arrays _now_ that would have enough angular resolution to pick out individual thunderstorms on the superjovian planets we've detected nearby. Resolving a beacon from a star spatially, for any star system near enough to matter, is do-able (though we aren't going to do it ourselves until we decide a space-based radio telescope array is worth the money).
I also question the parent post's assertations that radio signals are degraded to unintelligability. We can pick up millisecond pulsars just fine, meaning we could at the very least broadcast a beacon with data modulated at kHz rates. My understanding is that there are relatively clean frequency windows in the interstellar medium that would let us transmit intelligeably at far higher bandwidth.
That is a cool scenario. In our solar system, there is much extraterrestrial oxygen in the form of oxides, and relatively little in the way of flourides. Does that hold elsewhere? Who knows?
It turns out that oxygen is produced in great quantity both due to the CNO fusion cycle in massive stars, and nucleosynthesis during supernova explosions. So, I'd expect oxygen-dominated chemistry in most star systems.
Silicon is also a favoured nucleosynthesis product, which is why silicate rocks are so common.
OK, quick question. How would you be able to tell whether "our methods for establishing communication do not have a western or human bias"? What if the methodology to determine whether something is universal or culture?
Certain properties of the universe are independent of human existance. If I see a light shining at, say, exactly one third the wavelength of the 1->0 emission line of a hydrogen atom, I'll know it's blatantly artificial no matter what planet I evolved on.
As for communications methods, there appear to be only four fundamental forces, of which only two carry for great distances, and only one of which propagates in a way that lends itself to point-to-point communication. Unless we're very, very wrong about the nature of the universe, aliens will be using EM for their hypothetical beacons. This may not be radio, but there are limits to what it can be (certain bands don't carry well in the interstellar medium, and you get a high-energy cutoff due to the fact that information transmission requires a minimum photon count, and photons get more expensive to produce the higher the energy).
Chief, of course there cannot be any other beings across the Great Waters. If there were, we would have seen their smoke signals by now.
This falls into the "unless we're very wrong about the nature of the universe" category. The wonderful thing about this claim is that it's virtually impossible to _disprove_. However, I consider it unconvincing, because for the first time in human history, we have something approaching a _comprehensive_ model of reality. Certainly, there are extreme situations in which our models are known to not hold, but for everything else - from apples to transistors to stars - they do, in ways that are both testable and usable for things like engineering. This inspires confidence that our current understanding of reality is correct, at least for the domains where the models are intended to apply (important caveat).
In short, I think that EM transmissions are the right thing to be looking for, though SETI itself might not be searching for the right kind.
What's neutral to us may be either way beyond the comprehension of aliens (OK, so they're not intelligent), or so basic to them as to be meaningless (they're way, way, way more intelligent than we are).
It doesn't have to be profound - just something blatantly unnatural. The type of signal being broadcast would actually be a very clever way of setting a minimum standard for the intelligence of the creature you want to talk to on the receiving end.
A sequence of primes broadcast on radio waves would be picked up by any civilization as at mid-20th-century-earth advancement or higher. To send a message exclusively to the Elder Gods, you'd broadcst only concepts that they'd have reasonably found out about, optionally using hypothetical better communications methods that races like ours wouldn't know about.
Consider what it's like to explain prime numbers to bacteria... that's what it could be like for aliens to attempt "talking down" to us. Or, they could just see us as insignificant (but populous) little micro-organisms who just spread everywhere and reproduce in plague proportions with no other meaningful purpose.
Or more likely as serving some ecological niche, just as we view bacteria on Earth.
We don't know what alien civilizations will look like, and the spectrum of possibilities is vast. I see no problem with looking mainly for the kind that we'd be interested in talking to (the ones willing to speak in a way that we understand).
Pretty soon, if you don't know what you're looking for, this stuff becomes indistinguishable from gibberish.
[...]
So my question: Are there any standard texts or treatises on the theory of how to distinguish interesting signals from large amplitude noise?
Look at the entropy of the data stream, and look at the frequency spectrum of the signal. Most noise sources have very high entropy, and spectra that fall into a few well-defined shapes. Something intended to be a beacon or otherwise easily distinguishable from background noise would have low entropy (lots of redundancy), and a funny-looking spectrum (exact shape depends on modulation method, but it's easy to make something that looks blatantly artificial while still being easy to decode).
Signals (or rather, symbol streams in a decoded signal) that are intended to convey language have their own very-recognizable statistical patterns as well.
SETI assumes that the signals we're looking for are intended to be detected, and if detected, are intended to be easily interpreted - i.e., that they actually _are_ beacons of some kind, as opposed to leaked traffic. This makes detection much easier.
Leaked communications from a radio-using civilization with decent math and electronics skills is compressed to very high entropy and spread across as much spectrum as their transcievers can handle, resulting in something that's virtually indistinguishable from noise. Symbol stream statistics are similarly garbled. The only way to detect this kind of transmission is to look for big noise sources that aren't stars (Earth outshines the Sun in some radio bands, but resolving it separately from the sun at any distance requires an extremely large radio telescope array).
Even if chaos theory and quantum mechanics didn't exist and we could predict future events based on measurements of the the initial systems...why are you letting strangers come in and install complicated measuring equipment inside your random noise generator? Even if it was possible to know the initial state enough to predict the random numbers, which it isn't, how the hell would attackers manage to do it?
The idea is that they'd exploit knowledge of the system's workings to use its past output - which they could conceivably infer from your system's behavior - to derive the internal state, and from there attack the system. Alternatively, they could look at the state of the rest of the observable environment, and from knowing how that environment interacts with the RNG, either determine its internal state or drastically narrow down the possibilities. Or a combination of these.
If these sound like attacks on PRNGs to you, you're catching on. What a predictable universe does is turn any attempt at a true random number generator into a pseudo-random number generator, which can then be attacked using any of the standard approaches.
My point is that even in a deterministic universe, and especially in universes where you have both the uncertainty principle and nondeterminism to deal with, you can't produce meaningful predictions if the system is set up right, meaning you _can_ build what are for all intents and purposes true RNGs.
Whenever the debate between a random universe and a deterministic universe comes up, people always pull quantum mechanics out of their pocket and say, "Since we can never predict the future (Heisenberg), then the universe must be random." The inability predict the future has absolutely nothing to do with whether or not the universe is deterministic.
It does, however, mean that we can produce a RNG that is "random enough for all practical purposes", as we can never predict its outcome. That was the point of the original post.
That's why my first argument was made assuming a deterministic universe. Non-determinism isn't required (and was stated as a separate, third scenario component in my original post - the uncertainty principle was cited in the _second_ scenario to place further limits on what was knowable in the deterministic system).
Well, let's be clear here. Our best models (ie, quantum mechanics) have a statistical nature at their core. They are quite capable of predicting a large body of phenomena. Not all observed phenomena, and there's a few odd points here and there, but it's pretty damn good.
This does not, however, mean that this is the way the universe works.
That's why I put a great big "if" in front of "if quantum processes are accepted as truly random", and provided an argument showing that even if they aren't (e.g. if one of the hidden-variables models is correct, or what-have-you), the uncertainty principle limits what you can do in practice. Falling even farther back from that, chaos limits what you can do even in a perfectly deterministic universe where you can measure things to arbitrary precision, as there are parts of the past light-cone that you just plain can't see until you're past the measurement you're interested in, that have impact on the system.
So, given this, the "uncertainty principle" is the best we have. It's also somewhat axiomatic. It just is. But why? What gives rise to it at a very fundamental level?
It arises from the fact that any measurement requires an interaction, and any interaction perturbs the state of the system being measured in such a way as to destroy some of the information you're trying to measure. Look up the "gamma ray microscope" thought-experiment for a discussion of this.
To have this not hold, you'd have to revoke the wave/particle duality, which would cause a very large number of models to come crashing down (and more importantly, go against just about all experimental evidence to date).
Also, it hasn't been proved that it is possible to build such a machine yet. It may well be physically impossible to build a quantum computer of a useful size.
I'm placing my money on the "possible" side. There are plenty of engineering challenges, but no fundamental physics reasons it can't be done, and regular announcements of new techniques for overcoming the engineering problems. This is the hallmark of most new technologies, so I expect quantum computing to get very interesting as the engineering aspects begin to mature.
This is really quite simple - the type of machine that can render Prime-based and Discrete Log-based encryption "useless" has not been invented yet.
If by "prime-based" you're talking about deriving prime factors for things like RSA public keys, then the machines have been invented - they just haven't been built yet. Shor's Algorithm allows a quantum computer to factor numbers extremely rapidly, which breaks RSA quite badly. This is due to the nature of factoring, not of quantum computing itself - no quantum computing algorithm _presently_ known can break discrete-log encryption in less than the square root of the number of steps a classical computer would take to do it, for example. However, only time will reveal which algorithms are vulnerable to QC and which aren't.
Quantum computers with enough qubits to do useful RSA public-key factoring will probably be built within about 10-20 years, based on the progress to date. Possibly earlier, but I'm going to be conservative and hedge a bit.
It can well be argued that absolutely nothing is in fact random. From coin flips to roulette anything can eventually be learned and predicted on some level.
Even in a purely classical universe, sensitivity to starting conditions makes things like coin tosses and die rolls impossible to predict if set up carefully. This is that whole "chaos" topic you may have heard about in the press in the 1980s. You'd have to have excruciatingly accurate knowledge of the state of everything in the past light-cone of the event you're trying to predict, as of the time of prediction, for it to work with perfect reliability.
In our quantum universe, the uncertainty principle makes it impossible even in principle to measure starting state to the required precision, for the schemes that are used for true random number generation in electronic systems. Additionally, if quantum processes are accepted as truly random, they inject enough noise to taint macroscopic events with true randomness if the consequences of the noise are given enough time to propagate.
In summary, true randomness exists as a very fundamental result of the laws of nature, and won't go away no matter how good our measurements get.
Bzzzt. 90nM appears to be at the turning point where leakage becomes a significant part of total power.
There is a large difference between "significant" and "dominant". I am not disputing that it is a serious concern. However, it doesn't alter the validity of my response to the grandparent post.
We've passed the knee of the exponential curve on leakage current. 90nm leakage is worse by a few hundred percent compared to 0.13um.
And, at twice the clock speed, dynamic power dissipation is 200% worse. I'm not trying to claim that leakage isn't a problem - just that it tends to not be the dominant problem for high-speed chips. And yes, IAACE:).
As for dynamic power, don't ever expect wire capacitance to decrease. Even if you have thinner metal layer rules, remember that you've shrunk your area again, so the wires are also closer together and--let's face it--they are as long as ever on average.
Only for global routing. Local scales with line width, because you *really* don't want to deal with line delays when it can be avoided. Even long bus lines are segmented, for this reason, though that doesn't alter their capacitance.
It will take creative design to keep wattage in-check, but I don't expect it to level out. Clock for clock, 130 nm was worse than 180nm, and 90nm is worse than 130nm.
What ends up happening is that chip performance become power-bound instead of area-bound. To some extent, this is true already. What I expect to happen is a device technology switch along the lines of the ECL/CMOS switch that happened when ECL ran into the same problem years ago.
In the meantime, there are a number of interesting techniques used to reduce power dissipation. Adiabatic circuits are my personal favourite, though that's more because I think they're nifty than because I think they're a workable solution (current-density limits imposed by electromigration end up limiting your clock speed to something far lower than heating limits it to).
Both AMDs and Intel's projected offerings will be dual cache ("its easier that way") making them pretty much literally two seperate processors in a single package.
If I recall correctly the Power4 has a shared L3 (which they call "single-transistor SRAM", but which is actually DRAM backed by a number of SRAM buffers acting as cache for the cache).
I haven't checked on Sun's offering for a while, but IIRC it was supposed to have shared L3 as well. Not certain about that, however.
I wonder if somebody could explain why dual-core CPUs are a good idea. If it's a pair of cores on a single piece of silicon, it seems it would take the same silicon as two separate cpus, so where's the benefit?
Less packaging overhead, and faster communication between cores (on-die bandwidth and latency are far, far better than any motherboard's crossbar's bandwidth and latency).
You also have less contention over memory, for single-chip systems with multiple cores vs. multi-chip systems. Instead of having to muck about with cache coherence across a bus, the chip looks like a single processor as far as the memory subsystem is concerned, with coherence operations only involving the first one or two cache levels on-die.
yield decreases roughly exponentially with die size, which argues for 2 separate cpus.
Processes are optimized so that you can build a chip with 1-2 square centimetres of area with reasonable yield (as this is what chip manufacturers demand). This has been pretty constant (or if anything, has been increasing). However, with each design generation, the number of transistors available in this area has doubled. We're now at the point where we can get high yields on chips with enough transistors that multi-core designs make sense.
A chip with N cores also doesn't take N times as much area as a single-core chip, as the lowest levels of cache aren't duplicated (just L1 and usually now L2). So overhead is reasonable, and the real estate is there. It makes a lot of sense to use it.
In general, power dissapation scales in frequency with n^2, in multiple cores with n.
Power scales linearly with both frequency and the number of cores (or more accurately, with the amount of capacitance being switched per clock). It scales quadratically with _voltage_ (as capacitively stored energy is (1/2)CV^2).
Multi-core chips are used because we have more transistors available on a die, and both increasing cache size and increasing issue width on processors have reached diminishing returns for performance (though SMT-style chips help with the instruction issue problem). If you have twice as many transistors, the best way to improve performance nowadays is to build two cores on the die.
Slashdot Mirror
Cold fusion is impossible and Physics have long demostrated it.
Nothing is impossible.
Both of these statements are silly.
What we _can_ say about P&F "cold fusion" is that we have not been able, despite much trying, to produce convincing evidence of nuclear fusion in apparatus along the lines of the P&F "cold fusion" cells. The anomalous heat production almost certainly comes from something else.
As far as "nothing is impossible" is concerned, while from a philosophical perspective that statement is true, in practice, some physical laws (most of the ones taught in the textbooks that you ridicule, for instance) have been tested thoroughly enough that it's extremely unlikely they'll be violated in their domain of applicability. Step off a cliff, and you'd better have a bungee cord handy, because gravity isn't going to be repealed any time soon. Similarly, we've been studying nuclear effects for pretty much all of the 20th century. We know with confidence how excited nuclei drop to their ground states, and the fact that we haven't seen evidence of this in P&F "cold fusion" cells shows pretty conclusively that whatever's happening isn't fusion.
Anyways, that's my rant for today.
They didn't fabricate results, their results just became public too quickly
Oh? I seem to recall hearing about a neutron emission energy spectrum plot with a peak that kept wandering around between press conferences, until they finally withdrew it.
I'm going to have to pick up a copy of "Yes, We Have No Neutrons" one of these days so that I can have all of the questionable bits at my fingertips for situations like this.
I can't give information on the articles, but I can answer some of your questions:
1) Is that f not absorbed in space? What? Isn't space filled with H (relative to other elements).
Space is "filled with H" in that H is the least scarce element in space. What space is mostly filled with is a whole lot of nothing - laboratory-grade vacuum doesn't even come close to being as empty as most of space. So, radio signals (and light, for that matter) propagate vast distances without much trouble.
In the plane of the galactic disc, there's enough material that we can't actually see clear through the galaxy, but there are windows in the spectrum where absorption is much less, and that's probably where SETI is looking. Check the SETI page to see if they have details on the actual bands they're listening at.
2) How powerful must the xmiter be? That may tell alot.
This depends on how far away it is, how directional it is, how much noise your detector sees, and how you're trying to interpret the signal. A 1W source at a distance of 1 LY gives about 1e-33 W/m^2, if radiating in all directions. One radio photon has on the order of 1e-25 J (varies considerably with frequency). If you're listening with a dish with 1 m^2 area, you'll see an average one photon every 1e+8 seconds (3 years). That's how long you'd have to wait to determine whether the signal was present or absent in the complete absence of noise, which means you could pick up modulations of 1e-8 Hz at most, even under perfect conditions.
In practice, you have a noise floor which you have to rise above, and are looking for modulations in the MHz or greater range, and are looking at more distant sources, so power is much greater. Magnetic storms on stars are one of the main signal sources, for a sense of scale.
Of course, a walkie-talkie next to the dish, or an earthbound electrical storm anywhere nearby, would be picked up too. This is why many measurements are needed to verify a celestial emission source, and why radio telescopes are in remote areas.
3) what about com. by gravity? Is gravity instantaneous? though subject to r^2.
Gravity travels at the speed of light. This is why gravity waves exist.
5) We did signal back with the same signal followed by "hello" followed by the same signal. Didn't we? why not?
Because these telescopes are receivers, not transmitters, and because any response would take centuries or longer to reach its destination, and because the signal probably had an earth-based source (noise happens - a lot).
6) BTW, what would be the frequency signature of an H bomb reflected from a far away moon?
An h-bomb set off in space mainly gives off a pulse of hard radiation. If a moon is nearby, you might get x-ray backscatter from Bremsstrahlung radiation. If an object with an atmosphere is nearby, ionizing radiation striking the atmosphere may create lightning (as with very-powerful bombs set off on earth), which will give you radio noise that sounds much like any other electrical storm would (though it would differ by only being present briefly).
The flash and radiation pulse of the bomb itself would be brighter, but as it would have fewer photons, it might not be as easily detected far away. Either way, an H-bomb probably wouldn't be visible from another star system (just not enough photons produced). Communication between stars pretty much requires beamed communication, unless truly massive amounts of energy are involved.
I hope this information is useful to you.
As yet however, it is my understanding that every extra solar planet discovered has been by detecting the gravitational wobble rather than direct observation. As we have both said "in the future"(given enough money/time) it should be possible.
:)
I will begin to change my views once we actually *see* (radio or optical) them with our own eyes though
Running numbers, it looks like you'd only need a radio telescope array on the order of a few kilometres wide to resolve planets as distinct from their host stars within 100 LY or so. It seems odd that we haven't tried to directly view nearby superjovian planets this way.
My speculation is that it's because we'd need a very large filled-aperture dish to collect enough photons to sift the desired signal out of the background, but I'd be interested if any radio astronomers lurking could tell me the real constraints on trying this (as we know more or less where the target planets are).
I ought to know enough about discrete ergodic theory to know the answer to this, but is there a simple definition of "entropy" in this situation?
The short answer is "no"; it depends on how you decode the signal into a symbol stream, and several other parameters. However, most signals of natural origin have noiselike content no matter how you decode them, giving high entropy no matter how you measure it. An artificial signal that's intended to be easily picked up would be designed to have low entropy when interpreted in a very basic way (though this could be made more complex as an IQ test, per one of my other posts).
Similar problems with "the frequency spectrum" - just trying to decide on a time domain sampling interval [prior to applying a Fourier Transform] is a whale of a problem in and of itself.
The bandwidth and intensity of your signal impose limits on the data rates that may be present (due to photon count, and other environment-induced noise). This would let you pick reasonable parameters for frequency-domain analysis given the strength and bandwidth of the received signal. SETI's already doing something like this, I believe. The only major problem I can see offhand is that a signal may be modulated too _slowly_ to recognize with a small window, but I don't consider that a likely scenario if the signal is intended to be understood (they'd modulate it as quickly as possible while still having reasonable noise resistance).
As an aside, I'd ask you the same questions I asked another poster: Doesn't a Pulsating Radio Source [PULSAR] have a nice, stable, utterly utterly predictable spectrum? Similarly, doesn't 120V/60Hz [cf 240V/50Hz] wall current have a nice, stable, utterly predictable spectrum? Now which is interesting, and which is uninteresting? And does either one of them transmit a "message"? Or are they more akin to noise?
This question has confused astronomers as well - it was conjectured that pulsars were artificial beacons, until a natural explanation of them arose. An artificial signal intended to be picked up would be made unambiguous by encoding information that would not be generated by natural processes under any conditions. Encoding the first handful of prime numbers is a good way of doing this (either as binary, or as a series of delays between pulses, or what-have-you). Similarly, emitting a narrow-band signal at a small multiple or simple fraction of the wavelength of a naturally occurring spectral line would be recognizable as artificial. A repeating pulse is artificial-looking, but not artificial-looking enough to be unambiguous. So, I'd expect a beacon to have a more blatantly artificial pattern.
Most noise sources have very high entropy, and spectra that fall into a few well-defined shapes.
What are these "few well-defined shapes" [again, presumably relative to which "well-defined sampling intervals"]?
Flat spectra for white noise, 1/f spectra for "pink noise" or "flicker noise", and things like Poisson and normal distributions. Sampling interval doesn't really matter, here, so I'm not sure why you're bringing it up - think of this discussion as referring to continuous-time aperiodic transforms. Changing the scaling factor you're using alters the size but not the shape of the frequency spectrum that results.
Look up texts for semiconductor device physics for discussions of a number of examples of these types of noise come from. Celestial noise sources come from very different sources, but they still tend to follow the same types of distribution (as they arise naturally from the action of various simple classes of random processes - see a statistics text for a discussion of this).
The spectra of artificial signals is radically different, at least if produced by simple equipment or intended to be interpreted by simple receivers. You get a pair of spikes, lobes, or top-hats for a signal modulated onto a carrier. If you're deliberately making the signal as artificial-look
A chaotic system is *not* a random one, because, as you point out, given identical starting conditions, you get identical outputs. It's completely deterministic. I've still yet to see a convincing argument that randomness is more than a useful model for a set of deterministic processes we don't yet understand fully.
Look up "hidden variables" interpretations of quantum mechanics for discussions about this. The upshot is that while it's very difficult to prove the case one way or the other (and there are still people trying), you end up with a far cleaner model if quantum processes are treated as truly random. From there, in the absence of evidence of determinism, Occam's Razor kicks in.
I vaguely recall that there were actually experiments that ruled out some of the simpler "hidden variables" interpretations, strengthening the case for true randomness in quantum events, but you'll have to ask someone other than me for the details.
See, that's my problem. I'd like to go with your EM theory of communication, but I can't find any way of telling whether it's biased. So, how can I tell if what I've picked is subject to bias or not?
Please spell out how! That's all I'm asking.
The "can't possibly be biased" element from my original post was the frequency of the light being transmitted, not the use of EM in the first place. Hydrogen emission lines are fundamental constants of nature (at least in our neck of the woods; ObCaveats about very, very distant parts of the universe possibly having different values for these constants). Choosing a blatant derivative of one would be an unambiguous way of saying "look, I'm intelligent!" that is not tainted by human biases (except in the definition of "intelligence").
The argument for EM is that we seem to understand virtually all of the observable universe now, unlike your proposed witchcraft-using primitives, and EM looks like the _only_ choice for long-distance communication. The only possible objection is that we can't be sure that we understand the universe well enough to label alternate communications modes "unlikely".
I consider it likely that we understand the universe well enough to do this, as the relevant bits of physical theory have been around arguably even since Newton's day, and have stood despite very aggressive challenging (new models are best viewed as extending, not overturning, old, as they tend to reduce to the old models in the old models' domains of applicability). Your opinion may vary (and apparently does).
Right. Let's say we're broadcasting to a target audience of aliens at our level of civilization, say, between mid-20th-century and 10,000 years in the future (any more advanced, and they're not going to even bother... more important things to do).
So we're looking for aliens in a 10,000 year window. The universe is 15 billion years old, give or take. It sounds unlikely to me.
There are on the order of a hundred billion stars in our galaxy. With ObCaveats about assuming a uniform distribution across the 15 billion year domain, that still gives us an estimated 70,000 or so civilizations that are open to contact in our metaphorical backyard. We wouldn't get a timely answer from most of them, but my point is that they're still there, even with the numbers you postulate.
IMO, it's worth a shot. It costs us very little to look, and we learn something either way.
Dude, you totally dodged the question. Even after quoting it. Wow.
I'll spell it out for you, since you seem to have missed it, even after reading it:
The question of bias is irrelevant if you pick something that can't possibly be suceptible to bias. The universe is full of such things. Pick one.
Its not a problem of the actual signals degrading.
They get completely overpowered by the huge great big solar radio emitter, so that by the time they reach another starsystem, all thats resolvable is the signal from our sun itself.
This actually turns out not to be the case, for a couple of reasons. First, Earth outshines the sun on several radio bands - the sun's dumping most of its energy as visible light, and while electrical effects in its atmosphere are noisy, they don't cover the entire radio spectrum. Second, we could launch solar-orbit radio telescope arrays _now_ that would have enough angular resolution to pick out individual thunderstorms on the superjovian planets we've detected nearby. Resolving a beacon from a star spatially, for any star system near enough to matter, is do-able (though we aren't going to do it ourselves until we decide a space-based radio telescope array is worth the money).
I also question the parent post's assertations that radio signals are degraded to unintelligability. We can pick up millisecond pulsars just fine, meaning we could at the very least broadcast a beacon with data modulated at kHz rates. My understanding is that there are relatively clean frequency windows in the interstellar medium that would let us transmit intelligeably at far higher bandwidth.
That is a cool scenario. In our solar system, there is much extraterrestrial oxygen in the form of oxides, and relatively little in the way of flourides. Does that hold elsewhere? Who knows?
It turns out that oxygen is produced in great quantity both due to the CNO fusion cycle in massive stars, and nucleosynthesis during supernova explosions. So, I'd expect oxygen-dominated chemistry in most star systems.
Silicon is also a favoured nucleosynthesis product, which is why silicate rocks are so common.
OK, quick question. How would you be able to tell whether "our methods for establishing communication do not have a western or human bias"? What if the methodology to determine whether something is universal or culture?
Certain properties of the universe are independent of human existance. If I see a light shining at, say, exactly one third the wavelength of the 1->0 emission line of a hydrogen atom, I'll know it's blatantly artificial no matter what planet I evolved on.
As for communications methods, there appear to be only four fundamental forces, of which only two carry for great distances, and only one of which propagates in a way that lends itself to point-to-point communication. Unless we're very, very wrong about the nature of the universe, aliens will be using EM for their hypothetical beacons. This may not be radio, but there are limits to what it can be (certain bands don't carry well in the interstellar medium, and you get a high-energy cutoff due to the fact that information transmission requires a minimum photon count, and photons get more expensive to produce the higher the energy).
Chief, of course there cannot be any other beings across the Great Waters. If there were, we would have seen their smoke signals by now.
This falls into the "unless we're very wrong about the nature of the universe" category. The wonderful thing about this claim is that it's virtually impossible to _disprove_. However, I consider it unconvincing, because for the first time in human history, we have something approaching a _comprehensive_ model of reality. Certainly, there are extreme situations in which our models are known to not hold, but for everything else - from apples to transistors to stars - they do, in ways that are both testable and usable for things like engineering. This inspires confidence that our current understanding of reality is correct, at least for the domains where the models are intended to apply (important caveat).
In short, I think that EM transmissions are the right thing to be looking for, though SETI itself might not be searching for the right kind.
What's neutral to us may be either way beyond the comprehension of aliens (OK, so they're not intelligent), or so basic to them as to be meaningless (they're way, way, way more intelligent than we are).
It doesn't have to be profound - just something blatantly unnatural. The type of signal being broadcast would actually be a very clever way of setting a minimum standard for the intelligence of the creature you want to talk to on the receiving end.
A sequence of primes broadcast on radio waves would be picked up by any civilization as at mid-20th-century-earth advancement or higher. To send a message exclusively to the Elder Gods, you'd broadcst only concepts that they'd have reasonably found out about, optionally using hypothetical better communications methods that races like ours wouldn't know about.
Consider what it's like to explain prime numbers to bacteria... that's what it could be like for aliens to attempt "talking down" to us. Or, they could just see us as insignificant (but populous) little micro-organisms who just spread everywhere and reproduce in plague proportions with no other meaningful purpose.
Or more likely as serving some ecological niche, just as we view bacteria on Earth.
We don't know what alien civilizations will look like, and the spectrum of possibilities is vast. I see no problem with looking mainly for the kind that we'd be interested in talking to (the ones willing to speak in a way that we understand).
Pretty soon, if you don't know what you're looking for, this stuff becomes indistinguishable from gibberish.
[...]
So my question: Are there any standard texts or treatises on the theory of how to distinguish interesting signals from large amplitude noise?
Look at the entropy of the data stream, and look at the frequency spectrum of the signal. Most noise sources have very high entropy, and spectra that fall into a few well-defined shapes. Something intended to be a beacon or otherwise easily distinguishable from background noise would have low entropy (lots of redundancy), and a funny-looking spectrum (exact shape depends on modulation method, but it's easy to make something that looks blatantly artificial while still being easy to decode).
Signals (or rather, symbol streams in a decoded signal) that are intended to convey language have their own very-recognizable statistical patterns as well.
SETI assumes that the signals we're looking for are intended to be detected, and if detected, are intended to be easily interpreted - i.e., that they actually _are_ beacons of some kind, as opposed to leaked traffic. This makes detection much easier.
Leaked communications from a radio-using civilization with decent math and electronics skills is compressed to very high entropy and spread across as much spectrum as their transcievers can handle, resulting in something that's virtually indistinguishable from noise. Symbol stream statistics are similarly garbled. The only way to detect this kind of transmission is to look for big noise sources that aren't stars (Earth outshines the Sun in some radio bands, but resolving it separately from the sun at any distance requires an extremely large radio telescope array).
Even if chaos theory and quantum mechanics didn't exist and we could predict future events based on measurements of the the initial systems...why are you letting strangers come in and install complicated measuring equipment inside your random noise generator? Even if it was possible to know the initial state enough to predict the random numbers, which it isn't, how the hell would attackers manage to do it?
The idea is that they'd exploit knowledge of the system's workings to use its past output - which they could conceivably infer from your system's behavior - to derive the internal state, and from there attack the system. Alternatively, they could look at the state of the rest of the observable environment, and from knowing how that environment interacts with the RNG, either determine its internal state or drastically narrow down the possibilities. Or a combination of these.
If these sound like attacks on PRNGs to you, you're catching on. What a predictable universe does is turn any attempt at a true random number generator into a pseudo-random number generator, which can then be attacked using any of the standard approaches.
My point is that even in a deterministic universe, and especially in universes where you have both the uncertainty principle and nondeterminism to deal with, you can't produce meaningful predictions if the system is set up right, meaning you _can_ build what are for all intents and purposes true RNGs.
Whenever the debate between a random universe and a deterministic universe comes up, people always pull quantum mechanics out of their pocket and say, "Since we can never predict the future (Heisenberg), then the universe must be random." The inability predict the future has absolutely nothing to do with whether or not the universe is deterministic.
It does, however, mean that we can produce a RNG that is "random enough for all practical purposes", as we can never predict its outcome. That was the point of the original post.
That's why my first argument was made assuming a deterministic universe. Non-determinism isn't required (and was stated as a separate, third scenario component in my original post - the uncertainty principle was cited in the _second_ scenario to place further limits on what was knowable in the deterministic system).
Well, let's be clear here. Our best models (ie, quantum mechanics) have a statistical nature at their core. They are quite capable of predicting a large body of phenomena. Not all observed phenomena, and there's a few odd points here and there, but it's pretty damn good.
This does not, however, mean that this is the way the universe works.
That's why I put a great big "if" in front of "if quantum processes are accepted as truly random", and provided an argument showing that even if they aren't (e.g. if one of the hidden-variables models is correct, or what-have-you), the uncertainty principle limits what you can do in practice. Falling even farther back from that, chaos limits what you can do even in a perfectly deterministic universe where you can measure things to arbitrary precision, as there are parts of the past light-cone that you just plain can't see until you're past the measurement you're interested in, that have impact on the system.
So, given this, the "uncertainty principle" is the best we have. It's also somewhat axiomatic. It just is. But why? What gives rise to it at a very fundamental level?
It arises from the fact that any measurement requires an interaction, and any interaction perturbs the state of the system being measured in such a way as to destroy some of the information you're trying to measure. Look up the "gamma ray microscope" thought-experiment for a discussion of this.
To have this not hold, you'd have to revoke the wave/particle duality, which would cause a very large number of models to come crashing down (and more importantly, go against just about all experimental evidence to date).
Also, it hasn't been proved that it is possible to build such a machine yet. It may well be physically impossible to build a quantum computer of a useful size.
I'm placing my money on the "possible" side. There are plenty of engineering challenges, but no fundamental physics reasons it can't be done, and regular announcements of new techniques for overcoming the engineering problems. This is the hallmark of most new technologies, so I expect quantum computing to get very interesting as the engineering aspects begin to mature.
This is really quite simple - the type of machine that can render Prime-based and Discrete Log-based encryption "useless" has not been invented yet.
If by "prime-based" you're talking about deriving prime factors for things like RSA public keys, then the machines have been invented - they just haven't been built yet. Shor's Algorithm allows a quantum computer to factor numbers extremely rapidly, which breaks RSA quite badly. This is due to the nature of factoring, not of quantum computing itself - no quantum computing algorithm _presently_ known can break discrete-log encryption in less than the square root of the number of steps a classical computer would take to do it, for example. However, only time will reveal which algorithms are vulnerable to QC and which aren't.
Quantum computers with enough qubits to do useful RSA public-key factoring will probably be built within about 10-20 years, based on the progress to date. Possibly earlier, but I'm going to be conservative and hedge a bit.
It can well be argued that absolutely nothing is in fact random. From coin flips to roulette anything can eventually be learned and predicted on some level.
Even in a purely classical universe, sensitivity to starting conditions makes things like coin tosses and die rolls impossible to predict if set up carefully. This is that whole "chaos" topic you may have heard about in the press in the 1980s. You'd have to have excruciatingly accurate knowledge of the state of everything in the past light-cone of the event you're trying to predict, as of the time of prediction, for it to work with perfect reliability.
In our quantum universe, the uncertainty principle makes it impossible even in principle to measure starting state to the required precision, for the schemes that are used for true random number generation in electronic systems. Additionally, if quantum processes are accepted as truly random, they inject enough noise to taint macroscopic events with true randomness if the consequences of the noise are given enough time to propagate.
In summary, true randomness exists as a very fundamental result of the laws of nature, and won't go away no matter how good our measurements get.
Bzzzt. 90nM appears to be at the turning point where leakage becomes a significant part of total power.
There is a large difference between "significant" and "dominant". I am not disputing that it is a serious concern. However, it doesn't alter the validity of my response to the grandparent post.
We've passed the knee of the exponential curve on leakage current. 90nm leakage is worse by a few hundred percent compared to 0.13um.
:).
And, at twice the clock speed, dynamic power dissipation is 200% worse. I'm not trying to claim that leakage isn't a problem - just that it tends to not be the dominant problem for high-speed chips. And yes, IAACE
As for dynamic power, don't ever expect wire capacitance to decrease. Even if you have thinner metal layer rules, remember that you've shrunk your area again, so the wires are also closer together and--let's face it--they are as long as ever on average.
Only for global routing. Local scales with line width, because you *really* don't want to deal with line delays when it can be avoided. Even long bus lines are segmented, for this reason, though that doesn't alter their capacitance.
It will take creative design to keep wattage in-check, but I don't expect it to level out. Clock for clock, 130 nm was worse than 180nm, and 90nm is worse than 130nm.
What ends up happening is that chip performance become power-bound instead of area-bound. To some extent, this is true already. What I expect to happen is a device technology switch along the lines of the ECL/CMOS switch that happened when ECL ran into the same problem years ago.
In the meantime, there are a number of interesting techniques used to reduce power dissipation. Adiabatic circuits are my personal favourite, though that's more because I think they're nifty than because I think they're a workable solution (current-density limits imposed by electromigration end up limiting your clock speed to something far lower than heating limits it to).
Both AMDs and Intel's projected offerings will be dual cache ("its easier that way") making them pretty much literally two seperate processors in a single package.
If I recall correctly the Power4 has a shared L3 (which they call "single-transistor SRAM", but which is actually DRAM backed by a number of SRAM buffers acting as cache for the cache).
I haven't checked on Sun's offering for a while, but IIRC it was supposed to have shared L3 as well. Not certain about that, however.
I wonder if somebody could explain why dual-core CPUs are a good idea. If it's a pair of cores on a single piece of silicon, it seems it would take the same silicon as two separate cpus, so where's the benefit?
Less packaging overhead, and faster communication between cores (on-die bandwidth and latency are far, far better than any motherboard's crossbar's bandwidth and latency).
You also have less contention over memory, for single-chip systems with multiple cores vs. multi-chip systems. Instead of having to muck about with cache coherence across a bus, the chip looks like a single processor as far as the memory subsystem is concerned, with coherence operations only involving the first one or two cache levels on-die.
yield decreases roughly exponentially with die size, which argues for 2 separate cpus.
Processes are optimized so that you can build a chip with 1-2 square centimetres of area with reasonable yield (as this is what chip manufacturers demand). This has been pretty constant (or if anything, has been increasing). However, with each design generation, the number of transistors available in this area has doubled. We're now at the point where we can get high yields on chips with enough transistors that multi-core designs make sense.
A chip with N cores also doesn't take N times as much area as a single-core chip, as the lowest levels of cache aren't duplicated (just L1 and usually now L2). So overhead is reasonable, and the real estate is there. It makes a lot of sense to use it.
In general, power dissapation scales in frequency with n^2, in multiple cores with n.
Power scales linearly with both frequency and the number of cores (or more accurately, with the amount of capacitance being switched per clock). It scales quadratically with _voltage_ (as capacitively stored energy is (1/2)CV^2).
Multi-core chips are used because we have more transistors available on a die, and both increasing cache size and increasing issue width on processors have reached diminishing returns for performance (though SMT-style chips help with the instruction issue problem). If you have twice as many transistors, the best way to improve performance nowadays is to build two cores on the die.