Category: atomic clocks

Using Atomic Clocks for Time Synchronisation

  |   By

The atomic clock is unrivalled in its chronological accuracy. No other method of maintaining time comes close to the precision of an atomic clock. These ultra-precise devices can keep time for thousands of years without losing a second in drift – in comparison to electronic clocks, perhaps the next most accurate devices, which can drift up to a second a day.

Atomic clocks are not practical devices to have around though. They use advanced technologies such as super-coolant liquids, lasers and vacuums – they also require a team of skilled technicians to keep the clocks running.

Atomic clocks are deployed in some technologies. The Global Positioning System (GPS) relies on atomic clocks that operate onboard the unmanned orbiting satellites. These are crucial for working out accurate distances. Because of the speed of light that the signals travel, a one second inaccuracy in any GPS atomic clock would lead to positing information being out by thousands of kilometres – but the actual accuracy of GPS is within a few metres.

While these wholly accurate and precise instruments for measuring time are unparalleled and the expensive of running such devices is unobtainable to most people, synchronising your technology to an atomic clock, in actual fact, is relatively simple.

The atomic clocks onboard the GPS satellites are easily utilised to synchronise many technologies to. The signals that are used to provide positioning information can also be used as a source of atomic clock time.

The simplest way to receive these signals is to use a GPS NTP server (Network Time Protocol). These NTP servers use the atomic clock time signal from the GPS satellites as a reference time, the protocol NTP is then used to distribute this time around a network, checking each device with the GPS time and adjusting to ensure accuracy.

Entire computer networks can be synchronised to the GPS atomic clock time by using just one NTP GPS server, ensuring that all devices are within milliseconds of the same time.

The Hierarchy of a NTP Time Server Stratum Levels Explained

  |   By

When it comes to time synchronisation and using Network Time Protocol (NTP) to ensure accuracy on a computer network, it is important to understand the hierarchy of NTP and how it affects distance and accuracy.

NTP has a hierarchical structure known as stratum levels. In principle the lower the stratum number the closer the device is (in accuracy terms) to an original time source.

NTP time servers work by receiving a single time source and using this as a basis for all time on the network, however, a synchronised network will be only as accurate as the original time source and this is where stratum levels come in.

And atomic clock, either one sat in a large scale physics laboratory, or those aboard GPS satellites, are stratum 0 devices. In other words these are the devices that actually generate the time.

Stratum 1 devices are NTP time servers that get their source of time directly from these stratum 0 atomic clocks. Either by using a GPS receiver or a radio referenced NTP server, a stratum 1 device is as accurate as you can get without having your own multi-million dollar atomic clock in the server room. A stratum 1 NTP time server will typically be accurate to within a millisecond of the atomic clock time.

Stratum 2 devices are the next step down on stratum level chain. These are time servers that receive their time from a stratum 1 device. Most online time servers, for instance, are stratum 2 devices, getting their time from another NTP time server. Stratum 2 devices are obviously further away from the original time source and therefore are not quite as accurate.

The stratum levels on an NTP network continue on, with devices connecting to devices going all the way down to stratum 10, 11, 12 and so on – obviously the more links in the chain the less accurate the device will be.

Dedicated stratum 1 NTP time servers are by far the most accurate, reliable and secure method of synchronising a computer network and no business network should really be without one.

NTP and GPS-based Timing Solutions

  |   By

Ask anybody what the key to network timing is and you will probably get the response NTP (Network Time Protocol).  NTP is a protocol that distributes and checks the time on all network devices to a reference clock – and it is this reference which is the true key to network time synchronisation.

Whilst a version of NTP is easy to obtain – it is normally installed on most operating systems, or is otherwise free to download – but getting a source of time is where the true key to network time synchronisation lies.

Atomic clocks govern the global timescale UTC (Coordinated Universal Time) and it is this timescale that is best for network timing as synchronising all devices on a network to UTC is equivalent of having you network synchronised with every other UTC synced network on Earth.

So getting a source of UTC time is the true key to accurate network time synchronisation, so what are the options?

Internet Time Sources

The obvious choice for most NTP users, but internet time suffers from two major flaws; firstly, internet time operates through the firewall and is therefore fraught with security risks – if the time can get through your firewall, then other things can too. Secondly, internet time sources can be hit and miss with their accuracy.

Due to the fact most internet time sources are stratum 2 devices (they connect to another device that receives the UTC source time) and the distance from client to host can never be truly ascertained or accounted for – it can make them inaccurate – with some internet time sources minutes, hours and even days away from true UTC time.

Radio Referenced Time Server

Another source of UTC time which doesn’t suffer from either security or accuracy flaws is receiving the time from long wave radio signals that some country’s national physics laboratories broadcast. While these signals are available throughout the USA (courtesy of NIST) the UK (NPL) and several other European countries and can be picked up witha basic radio referenced NTP server they are not available everywhere and the signals can be difficult to receive in some urban locations or anywhere where there is electrical interference.

GPS-timing

For completely accurate, secure and a reliable source of UTC time there is no substitute for GPS time. GPS timing signals are beamed directly from atomic clocks onboard the GPS satellites (Global Positioning System) and received by GPS NTP time servers. These can then distribute the atomic clock time around the network.

GPS timing sources are accurate, secure and available literally anywhere on the planet 24 hours a day.

Time Synchronisation of Technology

  |   By

Many technologies are reliant and precise, accurate and reliable time. Time synchronisation is vital in many technical systems that we encounter everyday, from CCTV cameras and ATMs to air traffic control and telecommunication systems.

Without synchronisation and accuracy many of these technologies would become unreliable and in could cause major problems, even catastrophic ones in the case of air traffic controllers.

Precise time and synchronisation also plays an increasingly important part in modern computer networking, ensuring the network is secure, data is not lost, and the network can be debugged. Failing to ensure a network is synchronised properly can lead to many unexpected problems and security issues.

Ensuring accuracy

To ensure accuracy and precise time synchronisation modern technologies and computer networks the time controlling Network Time Protocol (NTP) is most commonly employed. NTP ensures all devices on a network, whether they are computers, routers, CCTV cameras or almost any other technology, are maintained at the exact same time as every other device on the network.

It works by using a single time source that it then distributes around the network, checking for drift, and correcting devices to ensure parity with the time source. It has many other features such as being able to assess errors and calculating the best time from multiple sources.

Obtaining the time

When using NTP, getting the most accurate source of time allows you to keep your network synchronised – not just together but also synchronised to every other device or network that uses that same time source.

A global timescale known as Coordinated Universal Time (UTC) is what most NTP servers and technologies use. A sit is a global timescale, and is not concerned with time zones and daylight saving, UTC allows networks across the world to communicate precisely with the exact same time source.

NTP time servers

Despite their being many sources of UTC across the internet, these are not recommended for accuracy and security reasons; to receive an accurate source of NTP there are really only two options: using a NTP time server that can receive radio transmissions from atomic clock laboratories or by using the time signals from GPS satellites.

NTP Servers Which Signal is Best Radio or GPS?

  |   By

NTP time servers (Network Time Protocol) are an essential aspect of any computer or technology network. So many applications require accurate timing information that failing to synchronize a network adequately and precisely can lead to all sorts of errors and problems – especially when communicating with other networks.

Accuracy, when it comes to time synchronization, means only one thing – atomic clocks. No other method of keeping time is as accurate or reliable as an atomic clock. In comparison to an electronic clock, such as a digital watch, which will lose up to a second a day – an atomic clock will remain accurate to a second over 100,000 years.

Atomic clocks are not something that can be housed in an average server room though; atomic clocks are very expensive, fragile and require full time technicians to control so are usually only found in large scale physics laboratories such as the ones run by NIST (National Institute of Standards and Time – USA) and NPL (National Physical Laboratory – UK).

Getting a source of accurate time from an atomic clock is relatively easy. For a secure and reliable source of atomic clock time there are only two options (the internet can neither be described as secure nor reliable as a source of time):

  • GPS time
  • UTC time broadcast on long-wave

GPS time, from the USA’s Global Positioning System, is a time stamp generated onboard the atomic clocks on the satellites. There is one distinct advantage about using GPS as a source of time: it is available anywhere on the planet.

All that is required to receive and utilise GPS time is a GPS time sever and antenna; a good clear view of the sky is also needed for an assured signal. Whilst not strictly UTC time (Coordinated Universal Time) being broadcast by GPS (UTC has had 17 leap seconds added to it since the satellites were launched) the timestamp included the information needed for NTP to convert it to the universal time standard.

UTC, however, is broadcast directly from physics laboratories and is available by using a radio referenced NTP server. These signals are not available everywhere but in the USA (the signal is known as WWVB) and most of Europe (MSF and DCF) are covered. These too are highly accurate atomic clock generated time sources and as both methods come from a secure source the computer network will remain secure.

The Time According to Cumbria Using the UKs MSF Time and Frequency Signal

  |   By

Getting an accurate source of time for computer networks and other technologies is increasingly becoming more important. As technologies advance and global communications mean that we are just as liable to communicate with technology across the other side of the planet as we are at home.

The need for accurate time is therefore essential if you wish to prevent time sensitive applications on your network failing or to avoid debugging problems – not too mention keeping your system secure.

NTP time servers (Network Time Protocol) are common devices that many computer networks use to provide a source of accurate time as NTP is able to ensure entire networks are synchronised to just a few milliseconds to the time reference.

The time reference that NTP servers use can come from several locations:

  • The internet
  • GPS satellite
  • And National Physical Laboratories

In the UK, the National Physical Laboratory (NPL) produce a time signal that can be received by radio referenced NTP time servers. This used to be broadcast from rugby in central England but in recent years the transmission has been moved to Cumbria.

The Cumbrian signal, known as MSF, is broadcast from Anthorn with a signal strength of 100 microvolts per metre at a distance of 1000 km. This should mean that the signal is available everywhere in the UK; however, this is not strictly the case as many MSF clocks and time servers can run into trouble when first trying to receive this atomic clock generated signal.

However, a simple checklist should ensure that no matter what your location you should be able to receive a signal to your MSF clock or NTP time server:

  • Check the power. Perhaps the most common problem ensure the battery is inserted and if the clock uses both mains power and a battery, remember to switch the mains power on. It can take quite a few minutes for the clock to pick up the MSF signal, so be patient.
  • Try rotating the clock or time server. As the MSF signal is long wave the antenna needs to be perpendicular to the signal for best reception.
  • If all else fails move the clock or time server to a different location. The signal can be blocked by local interference from electrical and mechanical devices.

* Note the MSF signal is down for scheduled maintenence on Tuesday 9 September 2010 from 10:00 BST to 14:00 BST

 

GPS Time Servers Precise Time all the Time

  |   By

Keeping computer networks accurate and synchronised can’t be emphasised highly enough. Accurate time is essential in the modern global economy as computer networks across the globe are required to constantly talk to each other.

Failing to ensure a network is accurate and precise can lead to headache after headache: transactions can fail, data can get lost, and error logging and debugging can be virtually impossible.

Atomic Clocks

Atomic clocks form the basis of the global timescale – UTC (Coordinated Universal Time). UTC is used across the globe by technology and computer networks enabling the entire commercial and technological world to communicate in synchronicity together.

But as atomic clocks are highly technical (and expensive) pieces of hardware that require a team of technicians to control – where do people get a source of such accurate time?

The answer is quite simple; atomic clock timestamps are transmitted by physics laboratories and are avlaible from a whole host of sources – kept accurate by the time software NTP (Network Time Protocol).

NTP Time Servers

The most common location for sources of atomic clock generated UTC is the internet. A whole host of online time servers are avlaible for synchronisation but these can vary in their accuracy and precision. Furthermore, using a source of internet time can create vulnerabilities in the network as the firewall has to allow these timestamps through and therefore can be utilised by viruses and malicious software.

By far the most secure and accurate method of receiving a source of atomic clock generated time is to utilise the GPS network (Global Positioning System).

GPS time servers are unique in that as long as there is a clear view of the sky they can receive a source of time – anywhere on the globe, 24 hours-a-day, 365 days a year.

They are also highly accurate with a single GPS NTP time server able to synchronise entire networks to just a few milliseconds of UTC.

How accurate does NTP Synchronisation need to be?

  |   By

Computers advance at a remarkable rate; in effect computers double in power, speed and memory every five years, and with such advances in technology many people assume that the clocks that control the time of a computer are just as powerful.

However, nothing could be further from the truth; most system clocks are crude crystal oscillators that are prone to drift, which is why computer time synchronisation is so important.

In modern computing, nearly every aspect of managing a network is reliant on time. Timestamps are the only frame of reference a computer has to ascertain if an event has occurred, is due to, or shouldn’t occur.

From debugging, to conducting time sensitive transactions over the internet, accurate time is essential. But how accurate does it have to be?

Coordinated Universal Time

Coordinated Universal Time (UTC) is a global timescale derived from atomic clocks. UTC was developed to allow technological devices, such as computer networks, to communicate with a single time.

Most computer networks use time servers governed by NTP (Network Time Protocol) to distribute UTC across the network. For most applications, accuracy to within a few hundred milliseconds is sufficient – but achieving this accuracy is where the difficulty lies.

Getting an accurate source of time

There are several options for synchronizing a network to UTC. Firstly, there is the internet. The internet is awash with time servers that proclaim to supply an accurate source of UTC. However, surveys of these online sources of time indicate that many of them are wholly inaccurate being seconds, minutes and even days out.

And even the most accurate and respected sources from NIST (National Institute of Standards and Time) and Microsoft, can vary depending on the distance your network is away.

Dedicated Time servers

Dedicated NTP time servers use a more direct approach to achieve accurate synchronisation. Using atomic clocks, either from the GPS satellite network or from physics laboratories (like NIST and the UKs NPL); the time is beamed directly to the NTP time server that is connected to the network.

Because dedicated devices like this receive the time directly from atomic clocks they are incredibly accurate, enabling the entire network to be synchronised to within just a few milliseconds of NTP.

Solar Flares and the Vulnerability of GPS

  |   By

Whilst GPS is commonly associated with satellite navigation and wayfinding, many technologies and computer networks rely on the GPS satellite system for a source of accurate time.

GPS time servers utilise the onboard atomic clocks of the global positing satellites and use this stable and accurate time source as a basis for their NTP synchronisation (Network Time Protocol)

GPS has become the preferred source of atomic clock time for many network operators. There are other methods where UTC (Coordinated Universal Time) can be used; radio signals and across the internet to name but two sources, but none is as secure or readily available as GPS.

Unlike radio signals, GPS is available everywhere on the planet, is never down for scheduled maintenance and is not commonly vulnerable to interference. It also doesn’t have any security implications like connecting across an internet firewall to an online time server can.

However, this doesn’t mean GPS is completely invulnerable as recent news reports have suggested.
It has been recently reported that a sunspot (sunspot 1092) the size of the Earth has flared up and a massive coronal ejection (solar flare), described in the press as a “solar tsunami” which was suggested to be large enough to satellites and wreck power and communications grids.

Solar activity, such as sunspots and solar flares (ejected hot plumes of plasma and radiation from the sun), have long been known to be able to damage and even disable satellites.

GPS is particularly vulnerable because of the high orbits of geostationary satellites (some 22,000 miles up) this leaves them unprotected by the earth’s magnetic field.

However, following the recent solar activity there has been no reported damage to the GPS system but as so many people rely on satellite navigation and GPS time for NTP servers could a future solar storm lead to havoc on Earth?

Well the short answer is yes; GPS satellites have been in orbit for several decades and while redundant satellites were introduced into the system many have been used up due to previous failures and it would only take a couple of disabled satellite to cause real problems for the network.

Fortunately, help is at hand as the Europeans, Russians and Chinese are all working on their own GPS equivalents which should work hand-in-hand with the American GPS network allowing GPS receivers to pick and choose from all four GNSS networks (Global Navigational Satellite Systems) ensuring that even if a really violent solar storm hits in the future there will be more than enough geo stationary satellites to ensure no loss of signal.

Using Internet Time for Computer Synchronization

  |   By

Ensuring your network is synchronized is a vital part of modern computing. Failure to do so, and having different machines telling different times is a recipe for disaster and can cause untold problems, not to mention making it almost impossible to debug or log errors.

And it is not just your own network you need to synchronize to either. With so many networks talking to each other, it is important that all networks synchronize to the same time-scale.

UTC (Coordinated Universal Time) is just such a global timescale. It is controlled by an international constellation of atomic clocks and enables computers all over the world to talk to each other in perfect synchronicity.

But how do you sync to UTC?

The internet is awash with sources of internet time. Most modern operating systems, especially in the Windows flavour, are set up to do this automatically (just by clicking the time/date tab on the clock menu). The computer will then regularly check the time server (usually at Microsoft or NIST, although others can be used) and adjust the computer to ensure its time matches.

Most internet time servers are known as stratum 2 devices. This means they take the time from another device but where does that get the time from?

NTP time servers

The answer is that somewhere on the stratum tree there will be a stratum 1 device. This will be a time server that receives the time direct from an atomic clock source. Often this is by GPS but there are radio referenced alternatives in several countries. These stratum 1 NTP (Network Time Protocol) time servers then provide the stratum 2 devices with the correct time – and its these devices we get our internet time from.

Drawbacks to Internet time

There are several drawbacks to relying on the Internet for time synchronisation. Accuracy is one consideration. Normally, a stratum 2 device will provide ample enough precision for most networks; however, for some users who require high levels of accuracy or deal in a lot of time sensitive transactions a stratum 2 time server may not be accurate enough.

Another problem with internet time servers is that they require an open port in the firewall. Keeping the NTP access on UDP port 123 open all the time could lead to security issues, especially as internet time sources can’t be authenticated or guaranteed.

Using a Stratum 1 NTP Time server

Stratum 1 NTP time servers are easily installed on most networks. Not only will they provide a higher accurate source of time but as they receive the time externally (from GPS or radio) they are highly secure and can’t be hijacked by malicious users or viral software.