The atomic clock is not a recent invention. Developed in the 1950’s, the traditional caesium based atomic clock has been providing us with accurate time for half a century.

The caesium atomic clock has become the foundation of our time – literally. The International System of Units (SI) define a second as a certain number of oscillations of the atom caesium and atomic clocks govern many of the technologies that we live with an use on a daily basis: The internet, satellite navigation, air traffic control and traffic lights to name but a few.

However, recent developments in optical quantum clocks that use single atoms of metals like aluminium or strontium are thousands of times more accurate than traditional atomic clocks. To put this in perspective, the best caesium atomic clock as used by institutes like NIST (National Institute for Standards and Time) or NPL (National Physical Laboratory) to govern the world’s global timescale UTC (Coordinated Universal Time), is accurate to within a second every 100 million years. However, these new quantum optical clocks are accurate to a second every 3.4 billion years – almost as long as the earth is old.

For most people, their only encounter with an atomic clock is receiving its time signal is a network time server or NTP device (Network Time Protocol) for the purposes of synchronising devices and networks and these atomic clock signals are generated using caesium clocks.

And until the world’s scientists can agreed on a single atom to replace caesium and a single clock design for keeping UTC, none of us will be able to take advantage of this incredible accuracy.

As with the advance of computer technology that seems to exponentially increase in capability every year, atomic clocks too seem to increase dramatically in their accuracy year on year.

Now, those pioneers of atomic clock technology, the US National Institute of Standards Time (NIST), have announced they have managed to produce an atomic clock with accuracy twice that of any clocks that have gone before.

The clock is based in a single aluminium atom and NIST claim it can remain accurate without losing a second in over 3.7 billion years (about the same length of time that life has existed Earth).

The previous most accurate clock was devised by the German Physikalisch-Technische Bundesanstalt (PTB) and was an optical clock based on a strontium atom and was accurate to a second in over a billion years. This new atomic clock by NIST is also an optical clock but is based on aluminium atoms, which according to NIST’s research with this clock, is far more accurate.

Optical clocks use lasers to hold atoms still and differ to the traditional atomic clocks used by computer networks using NTP servers (Network Time Protocol) and other technologies which are based on fountain clocks. Not only do these traditional fountain clocks use Caesium as their time keeping atom but instead of lasers they use super-cooled liquids and vacuums to control the atoms.

Thanks to work by NIST, PTB and the UK’s NPL (National Physical Laboratory) atomic clocks continue to advance exponentially, however, these new optical atomic clocks based on atoms like aluminium, mercury and strontium are a long way from being used as a basis for UTC (Coordinated Universal Time).

UTC is governed by a constellation of caesium fountain clocks that while still accurate to a second in 100,000 years are by far less precise than these optical clocks and are based on technology over fifty years old. And unfortunately until the world’s science community can agree on an atom and clock design to be used internationally, these precise atomic clocks will remain a play thing of the scientific community only.

Any network administrator will tell you how important time synchronization is for a modern computer network. Computers rely on the time for nearly everything, especially in today’s age of online trading and global communication where accuracy is essential.

Failing to ensure that computers are accurately synced together could lead to all manner of problems: data loss, security vulnerabilities, unable to conduct time sensitive transactions and difficulties debugging can all be caused by a lack of, or not adequate enough, time synchronization.

But ensuring every computer on a network has the exact same time is simple thanks to two technologies: the atomic clock and the NTP server (Network Time Protocol).

Atomic clocks are extremely accurate chronometers. They can keep time and not drift by as much of a second in thousands of years and it is this accuracy that has made possible technologies and applications such as satellite navigation, online trading and GPS.

Time synchronization for computer networks is controlled by the network time server, commonly referred to as the NTP server after the time synchronization protocol they use, Network Time Protocol.
When it comes to choosing a time server, there are really only two real type - the radio reference NTP time server and the GPS NTP time server.

Radio reference time servers receive the time from long wave transmission broadcast by physics laboratories like NIST in North America or NPL in the UK. These transmissions can often be picked up throughout the country of origin (and beyond) although local topography and interference from other electrical devices can interfere with the signal.

GPS time servers, on the other hand, use the satellite navigation signal transmitted from GPS satellites. The GPS transmissions are generated by atomic clocks onboard the satellites so they are a highly accurate source of time just like the atomic clock generated time broadcast by the physics laboratories.

Apart from the disadvantage of having to have a roof top antenna (GPS works by line of sight so a clear view of the sky is essential), GPS is obtainable literally everywhere on the planet.

As both types of time server can provide an accurate source of reliable time the decision of which type of time server should be based on the availability of long wave signals or whether it is possible to install a rooftop GPS antenna.

The Global Positioning System (GPS) is an increasingly popular tool, used throughout the world as a source of wayfinding and navigation. However, there is much more to the GPS network than just satellite navigation as the transmissions broadcast by the GPS satellites can also be used as a highly accurate source of time.

GPS satellites are actually just orbiting clocks as each one contains atomic clocks that generate a time signal. It is the time signal that is broadcast by the GPS satellites that satellite navigation receivers in cars and planes use to work out distance and position.

Positioning is only possible because thee time signals are so accurate. Vehicle sat navs for instance use the signals from four orbiting satellites and triangulate the information to work out the position. However, if there is just one second inaccuracy with one of the time signals then the positing information could be thousands of miles out – proving useless.

It is testament to the accuracy of atomic clocks used to generate GPS signals that currently a GPS receiver can work out its position on earth to within five metres.

Because GPS satellites are so accurate, they make an ideal source of time to synchronise a computer network to. Strictly speaking GPS time differs from the international timescale UTC (coordinated Universal Time) as UTC has had additional leap seconds added to it to ensure parity with the earth’s rotation meaning it is exactly 18 seconds ahead of GPS but is easily converted by NTP the time synchronisation protocol (Network Time Protocol).

GPS time servers receive the GPS time signal via a GPS antenna which has to be placed on the roof to receive the line of sight transmissions. Once the GPS signal is received the NTP GPS time server will distribute the signal to all devices on the NTP network and corrects any drift on individual machines.

GPS time servers are dedicated easy to use devices and can ensure millisecond accuracy to UTC without any of the security risks involved in using an internet time source.

We all rely on the time to keep our days scheduled. Wristwatches, wall clocks and even the DVD player all tell us the time but on occasion, this is not accurate enough, especially when time needs to be synchronized.

There are many technologies that require extremely accurate precision between systems, from satellite navigation to many internet applications, accurate time is becoming increasingly important.

However, achieving precision is not always straight forward, especially in modern computer networks. While all computer systems have inbuilt clocks, these are not accurate time pieces but standard crystal oscillators, the same technology used in other electronic clocks.

The problem with relying on system clocks like this is that they are prone to drift and on a network consisting of hundreds or thousands of machines, if the clocks are drifting at a different rate – chaos can soon ensue. Emails are received before they are sent and time critical applications fail.

Atomic clocks are the most accurate time pieces around but these are large scale laboratory tools and are impractical (and highly expensive) to be used by computer networks.

However, physics laboratories like the North American NIST (National Institute of Standards and Time) do have atomic clocks which they broadcast time signals from. These time signals can be used by computer networks for the purpose of synchronization.

In North America, the NIST broadcasted time code is called WWVB and is transmitted out of Boulder, Colorado on long wave at 60Hz. The time code contains the year, day, hour, minute, second, and as it is a source of UTC, any leap seconds that are added to ensure parity with the rotation of the Earth.

Receiving the WWVB signal and using it to synchronize a computer network is simple to do. Radio reference network time servers can receive this broadcast throughout North America and by using the protocol NTP (Network Time Protocol).

A dedicated NTP time server that can receive the WWVB signal can synchronize hundreds and even thousands of different devices to the WWVB signal ensuring each one is to within a few milliseconds of UTC.

Atomic clocks are the ultimate in timekeeping devices. Their accuracy is incredible as an atomic clock will not drift by as much as a second within a million years, and when this is compared to the next best chronometers, such as electronic clock that can drift by a second in a week, an atomic clock is incredibly more precise.

Atomic clocks are used the world over and are the heart of many modern technologies making capable a multitude of applications that we take for granted. Internet trading, satellite navigation, air traffic control and international banking are all industries that rely heavily on

They also govern the world’s timescale, UTC (Coordinated Universal Time) which is kept true by a constellation of these clocks (although UTC has to be adjusted to accommodate the slowing of the Earth’s spin by adding leap seconds).

Computer networks are often required to run synchronized to UTC. This synchronisation is vital in networks that conduct time sensitive transactions or require high levels of security.

A computer network without adequate time synchronization can cause many issues including:

Loss of data

• Difficulties in identifying and logging errors
• Increased risk of security breaches.
• Unable to conduct time sensitive transactions

For these reasons many computer networks have to be synchronized to a source of UTC and kept as accurate as possible. And although atomic clocks are large bulky devices kept in the confines of physics laboratories, using them as a source of time is incredibly simple.

Network Time Protocol (NTP) is a software protocol designed solely for the synchronisation of networks and computer systems and by using a dedicated NTP server the time from an atomic clock can be received by the time server and distributed around the network using NTP.

NTP servers use radio frequencies and more commonly the GPS satellite signals to receive the atomic clock timing signals which is then spread throughout the network with NTP regularly adjusting each device to ensure it is as accurate as possible.

Users of the National Physical Laboratory’s (NPL) MSF time and frequency signal are probably aware that the signal is occasionally taken off-air for scheduled maintenance.

NPL have published there scheduled maintenance for 2010 where the signal will be temporarily taken off-air. Usually the scheduled downtimes lasts for less than four hours but users need to be aware that while NPL and VT Communications, who service the antenna, make every effort to ensure the transmitter is off for a brief amount of time as possible, there can be delays.

And while NPL like to ensure all users of the MSF signal have advanced warning of possible outages, emergency repairs and other issues may lead to unscheduled outages. Any user receiving problems receiving the MSF signal should check the NPL website in case of unscheduled maintenance before contacting your time server vendor.

The dates and times of the scheduled maintenance periods for 2010 are as follows:

* 11 March 2010 from 10:00 UTC to 14:00 UTC

* 10 June 2010 from 10:00 BST to 14:00 BST (UTC + 1 hr)

* 9 September 2010 from 10:00 BST to 14:00 BST (UTC + 1 hr)

* 9 December 2010 from 10:00 UTC to 14:00 UTC

As these scheduled outages should take no longer than four hours, users of MSF referenced time servers should not notice any drop off in accuracy of their network as their shouldn’t be enough time for any device to drift.

However, for those users concerned about accuracy or require a NTP time server (Network Time Server) that doesn’t succumb to regular outages, they may wish to consider investing in a GPS time server.

GPS time servers receive the time from the orbiting navigational satellites. As these are available anywhere on the globe and the signals are never down for outages they can provide a constant accurate time signal (GPS time is not the same as UTC but is easily converted by NTP as it is exactly 17 seconds behind due to leap seconds being added to UTC and not GPS).

Accurate timekeeping if quite often overlooked as a priority for network administrators yet many are risking both security and data loss by not ensuring their networks are synchronised as precisely as possible.

Computers do have their own hardware clocks but these are often just simple electronic oscillators such as exist in digital watches and unfortunately these system clocks are prone to drift, often by as much as several seconds in a week.

Running different machines on a network that have different times – even by only a few seconds – can cause havoc as so many computer tasks rely on time. Time, in the form of timestamps, is the only reference computers use to distinguish between different events and failure to accurately synchronize a network can lead to all sorts of untold problems.

Here are some of the major reasons why your network should be synchronised using Network Time Protocol, prefasbly with a NTP time server.

Data Backups – vital to safeguard data in any business or organization, a lack of synchronisation can lead to not only back ups failing but older versions of files replacing more modern versions.

Malicious Attacks – no matter how secure a network, somebody, somewhere will eventually gain access to your network but without accurate synchronisation it may become impossible to discover what compromises have taken place and it will also give any unauthorised users extra time inside a network to wreak havoc.

Error logging – when faults occur, and they inevitably do, the system logs contain all the information to identify and correct problems. However, if the system logs are not synchronised it can sometimes be impossible to figure out what went wrong and when.

Online Trading – Buying and selling on the internet is now commonplace and in some businesses thousands of online transactions are conducted every second from seat reservation to buying of shares and a lack of accurate synchronisation can result in all sorts of errors in online trading such as items being bought or sold more than once.

Compliance and legality – Many industrial regulations systems require an auditable and accurate method of timing. A unsynchronised network will also be vulnerable to legal issues as the exact time an event is alleged to have taken place can not be proved.

When you counted down on New Year’s Eve to mark the beginning of the next year did you start at 10 or 11? Most revellers would have counted down from ten but they would have been premature this year as there was an extra second added to last year – the leap second.

Leap seconds are normally inserted once or twice a year (normally on New Year’s Eve and in June) to ensure the global timescale UTC (Coordinated Universal Time) coincides with the astronomical day.

Leap seconds have been used since UTC was first implemented and they are a direct result of our accuracy in timekeeping. The problem is that modern atomic clocks are far more accurate timekeeping devices than the earth itself. It was noticed when atomic clocks were first developed that the length of a day, once thought to be exactly 24 hours, varied.

The variations are caused by the Earth’s rotation which is affected by the moons gravity and tidal forces of the Earth, all of which minutely slow down the earth’s rotation.

This rotational slowing, while only miniscule, if it is not checked then the UTC day would soon drift into the astronomical night (albeit in several thousands of years).

The decision on whether a Leap Second is needed is the remit of the International Earth Rotation Service (IERS), however, Leap Seconds are not popular with everybody and they can cause potential problems when they are introduced.

UTC is used by NTP time servers (Network Time Protocol) as a time reference to synchronise computer networks and other technology and the disruption Leap seconds can cause is seen as not worth the hassle.

However, others, such as astronomers, say that failing to keep UTC in line with the astronomical day would make studying of the heavens nearly impossible.

The last leap second inserted before this one was in 2005 but there have been a total of 23 seconds added to UTC since 1972.

Oscillators have been essential in the development of clocks and chronology. Oscillators are just electronic circuitry that produces a repetitive electronic signal. Often crystals such as quartz are used to stabilise the frequency of the oscillation,

Oscillators are the primary technology behind electronic clocks. Digital watches and battery powered analogue clock are all controlled by an oscillating circuit usually containing a quartz crystal.

And while electronic clocks are many times more accurate than a mechanical clock, a quartz oscillator will still drift by a second or two each week.

Atomic clocks of course are far more accurate. They still, however, use oscillators, most commonly caesium or rubidium but they do so in a hyper fine state often frozen in liquid nitrogen or helium. These clocks in comparison to electronic clocks will not drift by a second in even a million years (and with the more modern atomic clocks 100 million years).

To utilise this chronological accuracy a network time server that uses NTP (Network Time Protocol) can be used to synchronise complete computer networks. NTP servers use a time signal from either GPS or long wave radio that comes direct from an atomic clock (in the case of GPS the time is generated in a clock onboard the GPS satellite).

NTP servers continually check this source of time and then adjust the devices on a network to match that time. In between polls (receiving the time source) a standard oscillator is used by the time server to keep time. Normally these oscillators are quartz but because the time server is in regular communication with the atomic clock say every minute or two, then the normal drift of a quartz oscillator is not a problem as a few minutes between polls would not lead to any measurable drift.

To be continued…