first lecture of the decade
Will surely be in 2021, not 2020...
The US-based GPS, a network of more than 30 satellites, is used by millions of phones, handsets and other devices in this country, for satellite mapping, navigation and communications technology. Then there's the European Union's Galileo system, which the UK had – until Brexit – contributed to building. Galileo, of course – …
While it's true the first lecture of the century was in 2001 and not 2000, since we're in the twenty-first century and not the "two thousands", since decades are described by their leading digits - i.e. we had the "nineteen-sixties", and not the "197th Decade", 2020 will be the first year of the twenty-twenties. 2021 will be the first year of the Two-Hundred and Third Decade, but no one will pay attention to that.
"Will surely be in 2021, not 2020..."
You can say that until you are blue in the face, and even implement it personally. However, the vernacular says otherwise. Unless you are willing, able and capable of convincing the GreatUnwashed of the wisdom of your words, all people will do is look at you funny and try to find someone else to sit next to at lunch.
A wise man doesn't fight unwinnable battles, Grasshopper ... and stop calling me Shirley.
I believe the Pope sides with the great unwashed. There's a story in which someone sells his soul to the Devil to be collected on the 1st day of the 20th century - the Devil being a mathematician duly turns up at midnight on 31st December 2000 only to be told that he is a year too late by Papal reckoning, and therefore the deal has lapsed.
From what little I know about atomic clocks, the typical caesium and rubidium clocks work off the microwave lines from hyperfine transitions. Electronics transitions allow use of light wavelengths and correspondingly higher precision. For higher precision yet one needs to cool the atoms. All this is established art.
What I'd like to know more about are what specific claims are made for these clocks and how they differ from the chip-scale rubidium references NIST is publishing on. Specifically, what's the root Allan variance? Settling time? Time transfer methodology?
The real neat trick is time transfer from device to device and handling the bookkeeping appropriately as one transfers time from stationary devices to moving devices such as aircraft. At caesium stability, relativity becomes apparent even for modest accelerations.
Mines the one with the 100 gram atomic clock in the pocket: https://www.orolia.com/products/atomic-clocks-oscillators
The real neat trick is time transfer from device to device and handling the bookkeeping appropriately as one transfers time from stationary devices to moving devices such as aircraft.
Not just moving devices - altitude too, so your aircraft was a good example. Time passes more slowly the closer you are to a centre of gravity. There was a TV programme a couple of years ago where they synchronised two portable atomic clocks, then took one up to the top of Snowdon and back. That one was then found to be ahead of the other because it had run faster when at altitude.
So if you really need the precision of atomic clocks, then you are going to have to allow for altitude.
"This is the key -- how much precision does your application really need ? I doubt that many need more precise than 1 ms."
Light travels at about 1 foot per nanosecond/1 meter per 3.34 ns, so if you wish to compute your location to the nearest foot/meter, that's how accurate your time needs to be.
Traditional GPS gives you your location to within 10m at a resolution of 40ns but I believe gains a little from multiple sources. 1ms gives you your location to the nearest 250 meters if my math is correct (probably not...)
So you can travel 300 km in 1ms at the speed of light. That is not the same as the accuracy of a location fix if your clock has a 1 ms skew.
My phone often has a skew of rather more than that on its clock and yet can get a decent satellite only location fix via GPS. Using ntp with say five reasonable sources will generally settle down to about 1 ms without its own stratum 0 source.
My phone often has a skew of rather more than that on its clock and yet can get a decent satellite only location fix via GPS.
Your phone doesn't need accurate time to use GPS (though it helps). But the GPS satellites, themselves need to have incredibly accurate time. So too, any time signal used for location calculation.
Welcome back GEE / LORAN-C
Given that, instead of putting portable atomic clocks in our smartphones at great expense, why not just put atomic clocks in our cell towers or something? In fact, once one gets an accurate position for every cell tower, it's quite possible that some old-fashioned method of radio navigation, not needing atomic clocks at all, could be used to find one's position without the help of satellites. Although maybe the atomic clocks would still help in rural areas.
If its only a a pulse that to be used in the determination then we go further back in time were at LORAN-A. LORAN-C used multi pulses to provide transmitter identification in each burst with phase comparison to complete the fix.
Now if the clocks are used to provide an ultra precise frequency, DECCA is back with lane counting.
Having worked at and on a Differential GPS site all I will honestly say is that the Velocity Factor varies in both the micro term and the long and any radio derived fix must receive the signals from several sources not local and with different frequencies. Omega was good but highly vulnerable to hostile intent.
"...My phone often has a skew of rather more than that on its clock and yet can get a decent satellite only location fix via GPS"
Light moves at about 300000 km/s. One millisecond corresponds to about 300 km, so I can guarantee you that the GPS circuitry in your phone has the time nailed down to better than a microsecond (300 meters). But I do know what you mean. I've seen devices where the time displayed clearly wasn't coming from the GPS.
On at least one older, stand-alone GPS unit, you could turn it on and watch the clock as it searched around for satellites. When it got a position fix, the clock would jump by a few seconds. It had a somewhat crummy quartz clock, just good enough to figure out which satellites might be visible. Once it could "see" four satellites, it had four equations and could solve for four parameters (x, y, z spatial coordinates and delta-T clock error). Then it would adjust the clock accordingly. (Four satellites are enough -- or three, plus a clock so you can set delta-T=0 -- but additional satellites do help you to reduce errors and give you some redundancy when you lose a satellite or two behind a building.)
Actually, with a single satellite you know that you can't be more than 15,000 km away from a satellite that you can connect to (satellites on the other side of the earth could be further away but they are invisible), so you know your time within 50ms. Don't think any GPS implements it :-)
Most people use GPS to find out what the time is, not to find out where they are.
If you are a bank and need to know what time someone withdrew some money, and whether it was before or after some other transaction; I don't think you need nanosecond accuracy, given that the process takes a few seconds to complete.
But they don't get better than microsecond accuracy and mostly do with millisecond accuracy. And some older systems still use whole seconds as most accurate level. And yes, I used to work as a programmer for a large, international bank and one of my jobs was to improve accuracy of timestamps to the microsecond level.
I doubt that many need more precise than 1 ms.
It's the *drift* that's more important. If the clock runs slow or fast by 1mS per day, it will be 1 second out in less than 3 years. Whether than is important depends on the application - especially how frequently it can be corrected from an accutate source.
"There was a TV programme a couple of years ago where they synchronised two portable atomic clocks, then took one up to the top of Snowdon and back"
i remember that tv show, was Jim Al-Khalili on beeb 4.
lots of different students took time devices to different places, up mountains, further north, south etc, was interesting.
not the beeb but think this might be it.
Hm. It derives its clock by exciting caesium molecules using a pair of lasers, though. Some sort of clever feedback system which keeps the laser excitation frequencies tuned very tightly to just either side of the critical resonance frequency of the molecules. So it is a genuine atomic clock.
I'm not explaining it well. Hang on...
It says "The SA.45s CSAC employs coherent population trapping (CPT) to interrogate an atomic frequency. A laser illuminates atoms in a resonance cell with polarized radiation at two sidebands separated by the atomic resonance frequency. The atoms are excited to a non-scattering coherent superposition state from which further scattering is suppressed. The small size and low power of the CSAC is enabled by a novel electronic architecture, in which much of the functionality of conventional atomic clocks has been implemented in firmware rather than hardware.
The SA.45s electronic hardware consists of a low-power digital-signal processor, a high-resolution microwave synthesizer, and analog signal processing. The microwave output is derived from a tunable crystal oscillator and is applied to the laser within the physics package to generate the two sidebands necessary for CPT interrogation. A photodetector detects light transmitted from the laser after it passes through the cesium vapor resonance cell. Based on the measured response of the atoms, the microprocessor adjusts the frequency of the crystal oscillator."
And then there's a whole lot of other stuff about how stable it is and why that is. It's all bloody clever, mind you. Still not accurate enough to calibrate the timing circuit of a Type 40 TARDIS though. For that you need beryllium.
I worked in a firm that became ISO-9000. A worthy cause, but some bureaucrats become a little too aggressive.
I had a caesium primary time reference in my electronics lab, adorned with a good half dozen ISO-compliant inventory tracking stickers, but no Cal Lab sticker.
Cal lab: "You have a piece of equipment we need to cal! And you refuse to send it to us! Waahhh!"
Me: "It's ... A primary standard. You, uh, should know already that by definition we Cal other crap off it IT! Given that you don't grasp that, I'm ... concerned."
Cal lab: "Bad boy! It has stickers on it. That means it can be cal'ed! You're just being a poopyhead!"
Yeah! Turtles, er I mean documents, all the way down! When your team starts stamping each other's foreheads with property management stamps before the inspectors come in, and labelled every toilet in the men's room "This container is not authorized for the storage of sensitive information" ... and "report any information spillage to your security officer!" on the floors under the urinals ... you're approaching compliance.
One of the more "interesting" parts about using Loran for nav is that you need to know what KIND of geography is around the area.
Radio diffracts/refracts quite differently around a pointy peninsula than a round-ended one and differently still along a smoothish coastline - this would frequently result in Loran-derived positions that were several miles from where you actually were - not good if the area in question has rocky reefs in it (or if you're giving Loran positions for SAR purposes - it wasn't uncommon for them to be 20 miles out and a small boat in open ocean in bad conditions is remarkably hard to spot even when only 5 miles away in the air)
There's a very good reason it fell out of favour quickly when civilian GPS got into high-accuracy mode.
A very small atomic clock will be a significant technical achievement but I can't see how it provides a means of navigation or removes the need for GPS.
It could be part of a navigation system as atomic clocks already ready are in satellite navigation systems. On the face of it this is just hype based on the first thing that has any connection to precision time keeping, or I am being very stupid and there is someway that a really accurate time source provides position without the need for signals from satellites, measuring the position of astronomical objects etc.
The hype seems to originate from the University of Sussex or the researchers as it is not just in the register.
I was just about to ask the same question.
I always thought that the whole purpose of having the signals beamed from satellites was that you could triangulate from the three known coordinates, and the only reason for the clock signals was to measure the tiny variances in the time taken for the signals to reach the receiver.
So how does having your own portable clock absolve the need for the satellite signals? How doe the unit triangulate?
But how does your device know the distance to each satellite?
If only there was something that could be measured that when combined with an initial location (i.e. latitude/longitude/altitude) and a known speed (close enough to the speed of light plus some adjustments for atmospheric effects) that would allow you to work out distance.
But how does your device know the distance to each satellite?
It doesn't (initially). The calculation is very much chicken-and-egg. Initially you only know the approximate *difference* in distance between you and each satellite being received - this being calculated from the difference in the time-stamps each satellite sends from its on-board atomic clock. (It is approxiate because the amount of refraction of each signal is at that point unknown). The latest time stamp being from the satellite closest to you, and the earliest time-stamp from the most distant satellite. As you do not know the absolute distance between you and any satellite, you also do not know the real time to any great accuracy, and hence the exact position of each satellite in its orbit.
But so long as the geometary (relative positions) of the satellites are favourable, then there will only be one place on the surface of the Earth where any particular combination of distance differences between 3 or more satellites can occur. So if you are receiving 3 or more suitably placed satellites you can calculate your approximate position. Then you can reverse calculate the distance to any satellite, also its elevation to get a more accurate figure for the amount of refraction of its signal as it passes through the ionosphere, and from that calculate the time its signal took to reach you, and hence get a more accurate real-time as given by the satellite's atomic clock.
So the whole thing is re-calculated with the more accurate information. Do this several times (iterations) and the position will (should) converge until it stops changing between iterations or reaches the desired accuracy.
Note that the refraction amount is not an accurate variable even if you know your time & position 100% because the ionosphere is constantly changing in strength, height and thickness in a way that cannot be accurately determined, and so the amount of refraction cannot be known with absolute accuracy. This puts a limit on the best positional accuracy possible with the system, especially if all the satellites used are low to the horizon.
If you have an accurate real-time clock, then you could calculate position with only 2 satellites rather than 3 (Add one satellite if you are not on the surface of the Earth).
I'm afraid you explanation is incorrect. A 3d fix needs 4 satellites as you need to solve for time and 3 spacial variables, which is why GPS receivers can typically provide very accurate time output, as they already solve for that.
The time/frequency references used for static telecoms start with a survey mode which removes 3 variables, and then allows you to get pretty accurate time with just one satellite (the ephemerides and an L2 receiver can give you a lot of ionospheric compensation).
Portable atomic clocks are great tech, but won't really help with GPS as even if you ignore relativistic drifts, you'd still need 3 sats for a 3d fix.
The reason you need an accurate clock is so that you know exactly where the satellites are in their orbit. The difference in the times reported by the satellites gives you their relative distance from each other to you (the signal takes longer to travel from the furthest satellite). As this only gives you relative distance compared to all other satellites being received, you need at least 3 satellites to compute a position (so long as you know your altitude). If you had an accurate local clock, you would only need 2 satellites.
The old "transit" satellite navigation system used the change in doppler shift of 1 satellite as it orbitted to find a position, but that depended on knowing the distance and direction you had moved while taking the readings (which took about 20 minutes to complete). A transit satellite would pass within range about every 2 hours on average (more frequently the closer you are to the poles). So you could only obtain a position about 12 times per day. So no good for precision real-time navigation.
You don't need a local accurate clock: with 3 satellites you do, but with 4 you can solve for position and time. All that this would do is eliminate the need for one degree of the solution, but you really want more than the minimum anyway because (a) some might close together leading to the "geometric dilution of precision" and poorer fix accuracy and (b) with N+1 you can tell if one satellite is flaky, and with N+2 you can tell which one it is.
So really all that you get from an accurate local clock is a benefit if you are using GPS for time sync (important for mobile phones, DTV, etc) as you can run for longer between synchronisation times. What you don't get is navigation without some form of triangulation (even if star tracking...)
If you also have accurate accelerometers you could dead reckon.
You would still need the occasional* reference, perhaps from land based systems
*how occasional depends on the accuracy of your clock and accelerometers.
I believe inertial navigation is still one input to commercial aircraft navigation systems
I'm quite sure I have previously read about ship navigation beacons similar to how a lighthouse works. The beacon ( lighthouse would be a good place to situate it) sends a very accurate time signal along with a location ID, some clever maths and a lookup to the location database would be able to give a very accurate distance from the known location of the beacon. If you have 2 or more beacons in range you can calculate your position, the more beacons in range more accurate you can be. Virtually the same as GPS and other space based versions but obviously needs more beacons due to the curvature of the earth etc (if its flat.. I'm not going there).
Anyway the ground based version should be massively cheaper for local navigation / busy shipping lanes where line of sight is not such an issue.
I think the main problem would be putting up these beacons in enemy territory so bombs could be dropped where you intend.... And that's why GPS etc can be turned off and has encrypted channels, just in case you know.. Oceana are allies with Eastasia today...
"These were land radio based systems that could provide a position fix to within 300m or so"
Supposedly. Under ideal conditions. Close to the transmitters
Reality, earth curvature, ionospheric waveguide effects, time of day and the shape of the land/sea interface always degraded that somewhere between slightly and "you won't believe how far out it is"
As an example, bank ATMs have GPS receivers in them.
One use of that probably will be to make sure that the ATM is still positioned in the wall of the bank, and to do something if it detects movements; however, the most important use of it is simply to get a precise time signal, because it needs to know exactly what time things happen.
This is actually the most common use of GPS - to get the time rather than to get location.
All bank ATMs have a direct connection (secured internet) with a server of that bank and can use the time server of that bank that way. The time server of the bank just polls one or more NTP servers, no need whatsoever to complicate things with GPS, which signal may not even be available inside a building.
... someway that a really accurate time source provides position without the need for signals from satellites, measuring the position of astronomical objects etc.
Yes. It's called "celestial navigation," discovered (invented?) in 1837 by Thomas Hubbard Sumner. The big problem at the time was that there was no way to determine the time with sufficient accuracy when you did not know your position - thus prompting the race to invent a super-accurate clock that was impervious to both a ship's movement and temperature changes.
" It's a system that employs portable, optical atomic clocks as a means of extremely accurate time-keeping that theoretically completely ends the need for navigation systems here on Earth to talk to an orbiting network of satellites."
My phone TALKS to an orbiting network of satellites? And there was me thinking it just listened.
"Synchronise watches... I make it 12:23:57... NOW!!" "Sorry sir, I make it only 12:23:56.999999999999999999999999999999999999"
When we all have our personal atomic clocks, who sets the standard?
Hey, why not have a cluster of master clocks that we could all synchronise our pocket atomic clocks to..
And even better/easier/cheaper, we could do away with having our own portable atomic clocks and just synchronise all our devices to it, just like, erm...
> the European Union's Galileo system, which the UK had – until Brexit – contributed to building .. remains accessible to the UK after Brexit, although the country will cut itself off from access to its encrypted channel after the country's departure from the union.
I don't understand the logic of rescinding full GB access to Galileo after Brexit, as long as they continued to provide finance and technical support, after all a lot of the technology was provided by private contractors and it's not as if the Island is going to disappear into orbit.
> The EU says the agreement struck over the terms of the 21-month transition period after Brexit gives it the freedom to do so ..
Why would GB agree to be blocked from Galileo post Brexit, unless this is yet another example of Brussels making up the rules as they go along.
"The commission has suggested" you turn over your Laundry Service to our friend Whispers :]
>As I recall, it was the UK who insisted that that particular bit of regulation was included.
Regulations can be changed,
That surely is the whole point of working within a sensible, democratic, efficient and technocratic organisation such as the Eu.
Neither side would retreat to childish blocking of cooperation for public point scoring or to further political gains in other areas
If the other countries in the EU change the rules to suit us whenever we want then they surely must make up a sensible, democratic, efficient, and technocratic organisation and they are good.
Otherwise if they don't change the rules to suit us whenever we want then they are crazy, antidemocratic, burocratic, inefficient, and, er, technocratic and they are therefore evil.
Thanks to the likes of Farage and the current bunch of EMPs, any sympathy for the UK has long disappeared in the EU. The UK insisted that things would be only available to EU members. And the UK insisted on leaving the EU. Not giving anything to the UK isn't for public point scoring or political gains, but just to tell them to feck off.
> As I recall, it was the UK who insisted that that particular bit of regulation was included.
Do you have a link to the actual text where GB agreed to be excluded from Galileo post Brexit? According to that Guardian quote, the EU assumed the freedom to do so, but that is a curious phrase to use in relation to a legal agreement. Under that criteria, if you sold me your furniture, I could assume the rest of the house came with it. If you objected I could merely say but you gave me the freedom.
"As I recall, it was the UK who insisted that that particular bit of regulation was included."
At the absolute insistence of the USA
Who wanted to ensure China was locked out of Galilleo
Against the objections of every other member.
So when Britain announced it was leaving, there was ZERO thought of changing the rules so the UK could stay.
You know that saying about being careful what you wish for?
(this isn't the only example of the UK forcing rules on the rest of the EU which are now being used against it as it's leaving, whilst the rest of the EU smirks and says "sorry, we can't change the rules")
>Considering the UK Government is now run from the Kremlin
The current occupant of No 10 is an Eton and Oxford educated classicist with no obvious skills but who has mysteriously risen to power - hardly the traditional profile of a Russian spy within the British establishment
One problem with GPS, GLONASS and Beidou is that they are all primarily military systems, under the control of the military leadership in their respective countries and their commercial services can be turned off at will, for military reasons, in part or all of their operating areas. Galileo is not a military system and one of the primary intended uses is for transport, in conjunction with EGNOS. In particular, EGNOS and Galileo will allow aircraft full CAT1 Autloand in a few years time, without the need for ILS infrastructure at airports.
Barring non-EU countries from PRS use was a British idea but the EC has acted somewhat vindictively where it comes to the UK. Last spring, the EC sent a memo to all EU27 national governments, telling them that "the UK is not to be trusted on matters of security". This was at the same time as the EC were insisting that they have access to UK intelligence information, as part of the withdrawal agreement and when UK forces have been protecting eastern Europe from the big bad Russians, under the auspices of NATO. The EC's attitude and complications as regards non-NATO EU members maybe encouraging an increase in the importance/use of 5 Eyes.
The EC's attitude and complications as regards non-NATO EU members maybe encouraging an increase in the importance/use of 5 Eyes.
This is pretty normal, though. Complications about who helps whom at the national / territory / treaty / bloc / etc level are always going to be a thing. Military types are used to it, and very good at it.
The tricky thing is that the EC has taken this a bit further: They have offered that the UK could use Galileo PRS - but only in activities that were coordinated with the EC or in activities that were inline with the EU's policies/interests. The implication of the EC's offer is that they would reserve the right to deny the use of PRS to the UK at any time, if they felt it was not in their interest or simply that they did not approve.
Every time something like this is said, it devalues the lecturer in question. They have a good contribution to make, can communicate well and are very smart; those are the criteria.
Monitoring (and praise) of any other attributes will always start to imply (or in the worst case, create) a culture where people have been picked for said other attributes, because if there aren't enough good people who just happen to have those attributes, then less good people will be picked just because of those attributes.
For navigating in your home waters they are rather excellent, although a trifle heavy for carrying while mountaineering.
For dropping precision guided munitions on the fuzzy-wuzzys they are are bit limiting.
Once you have successfully negotiated with the recalcitrant natives to install your navigation beacons it seems rater bad form to then bomb them.
A whispering gallery to get the exact location for slide windows?
I suppose any sufficiently accurate independent time source would work though.
Incidentally I finally managed to source an FE-5680 but need to find out if using an SDR directly would avoid the need for all that bulky and heavy circuitry from the early Netscape era. This is possible as the actual core which gets to about 100 celsius when in use comparable to a laptop CPU will fit into a far smaller case assuming its not needed all the time ie just for calibration purposes.
For calibration it does not need to output 10 MHz as long as I know what the correct frequency is and add external filters accordingly.
The nice folks at some 'Uni invented an optical and electrostatic trap that "locks" Sr ions.
In principle this system would be ideal for something like this.
I looked into DIYing a Rb clock which is possible thanks to advances in SDR and lasers: the LNBs from a few years back will with some modification tune into the Rb hyperfine transition at 6.8 GHz so a bit of RbCl2 off eBay and some argon eg from a new fluorescent tube starter with the SRD made from a modified tunnel diode or used GaAs based Gunn out of a broken door scanner should work.
You could literally DIY an atomic clock this way but it would be a lot of work comparable with a doctoral thesis and making a double oven OCXO using parts from a broken 8088 MB is possible.
Cs is harder but the transition is closer to the range on LNB so you could do the same here however it will never be as good as a proper CSAC or used module with regenerated lamp.
Why the thumb down? Also my double oven OCXO showed up.
The mini £30 Rb lamps are actually quite easy to use but do have issues: seems that some have off centre pinches and other manufacturing glitches so might need some creative engineering.
This is still a good idea even if it takes more signal processing than I first thought.
Note: Making a SRD from a broken Gunn oscillator is literally nearly impossible but you might luck out. Negative resistance diodes do often run at strange harmonics and the mixer diode is even wider range.
https://pic-microcontroller.com/rubidium-atomic-clock/ (Not mine!)
The argon gas actually increases the frequency slightly but as its a known effect it can be compensated for, also 3D printed silver cavity is feasible with some post printing fine etching to get the exact shape.