Time Synchronisation

This article explores how to use national time and frequency radio transmissions for network time synchronisation.

The importance of an authenticated timing reference to synchronise a computer network to, cannot be stressed highly enough.

While there are hundreds and quite possibly thousands of internet based timing sources these can’t be authenticated leaving a system open to viruses, malicious hackers or malware.

Furthermore, a survey by MIT (Massachusetts Institute of Technology) found that nearly half of internet timing sources were offset by over ten seconds and only a third could be regarded as being ‘useful ’ also it was discovered that many were too far away from peers to provide any useful accuracy.

Most dedicated network time servers are designed to receive a timing signal from the GPS (Global Positioning System), primarily because it is the most accurate and can be received from anywhere on the globe.

However, there are situations where it may not be practical to use a GPS time server. A GPS antenna has to be situated on a rooftop and have a clear view of the sky which may prove difficult if the server is on the ground floor of a multi-storey sky-scraper. Many administrators also dislike the hassle and expense of having to run a cable up a building and install an antenna or if there are possibilities the server room maybe relocated and the process has to be repeated.

Fortunately many countries’ national physics laboratories broadcast a time and frequency signal from a radio transmitter. In the US the signal is referred to as WWVB and is broadcast by NIST (National Institute for Standards and Technology) in Colorado. In the UK the National Physical Laboratory (NPL) broadcasts the MSF signal from Cumbria and similar systems are broadcast in Germany (DCF-77), Japan (JJY) and France (TDF).

Unfortunately not every country transmits a national time and frequency broadcast so if a time server is to be located outside of the US, Germany, UK, France or Japan it may be doubtful that a signal could be received (although many of the these transmissions can be received in neighbouring countries).

Radio signals are also easily susceptible to atmospheric interference and can be blocked by mountains, sky-scrapers or other topography. However, an upside to using a radio receiver is that it will receive a signal inside a building.

While a radio transmission is not as accurate as a GPS time signal a dedicated network time server receiving a radio signal can still provide accuracy between 1 – 20 milliseconds (a millisecond is 1/1000 of a second) which is more than adequate for the needs of network synchronisation.

Many people are probably familiar with the Internet Time tab when setting their clock in Windows. This is a basic form of NTP (Network Time Protocol) called SNTP (Simple Network Time Protocol) that polls a NTP server every so often to synchronize time to. However, full NTP does a lot more, such as polling several servers to determine what is the best and most stable times source.

Creating your own NTP time server is relatively straightforward and if you are a resident of the US then the simplest way is to use a relatively cheap receiver module, set to receive the WWVB timing signal which is broadcast by NIST (National Institute of Standards and Time) at a frequency of 60 kHz.

To create your own product using the WWVB time signal, WWVB receiver modules, are readily available at low cost. When looking for a receiver module there are several points to consider:

There should be a simple interface to allow easy integration of accurate timekeeping into electronic equipment etc. The receiver should use dedicated chips designed specifically to receive the WWVB (USA) time signal. The advantage of these over other solutions is that the modules include the necessary support electronics, tuned crystals, etc and are preassembled, tested and are usually in miniature form.

When combined with a suitable Antenna the receiver module acts as a complete time code receiver, providing a serial digital data output stream for external decoding.

Once assembled and tested so the module is receiving a signal it can then be connected to your computer or server.

Once connected the time server should be configured using the registry editor. Simply follow these steps:

Locate the following subkey: HKEY_LOCAL_MACHINESYSTEMCurrentControlSetServicesW32TimeParametersType
In the right pane, right-click Type then click Modify, in edit Value type NTP in the Value data box then click OK.

Locate the following subkey: HKEY_LOCAL_MACHINESYSTEMCurrentControlSetServicesW32TimeConfigAnnounceFlags.

In the right pane, right-click AnnounceFlags and click Modify. The ‘AnnounceFlags’ registry entry indicates whether the server is a trusted time reference, 5 indicates a trusted source so in the Edit DWORD Value box, under Value Data, type 5, then click OK.

To enable the Network Time Protocol; NTPserver, locate and click: HKEY_LOCAL_MACHINESYSTEMCurrentControlSetServicesW32TimeTimeProvidersNtpServer

In the right pane, right-click Enabled, then click Modify. In the Edit DWord Value box, type 1 under Value data, then click OK.

Now go back and click on: HKEY_LOCAL_MACHINESYSTEMCurrentControlSetServicesW32TimeParametersNtpServer

In the right pane, right-click NtpServer, then Modify, in the Edit DWORD Value under Value Data type In the right pane, right-click NtpServer, then Modify, in the Edit DWORD Value under Value Data type the Domain Name System (DNS), each DNS must be unique and you must append 0×1 to the end of each DNS name otherwise changes will not take effect.

Now click Ok then locate and click the following: HKEY_LOCAL_MACHINESYSTEMCurrentControlSetServicesW32TimeTimeProvidersNtpClientSpecialPollInterval

In the right pane, right-click SpecialPollInterval, then click Modify. In the Edit DWORD Value box, under Value Data, type the number of seconds you want for each poll, ie 900 will poll every 15 minutes, then click OK.

To configure the time correction settings, locate: HKEY_LOCAL_MACHINESYSTEMCurrentControlSetServicesW32Timeconfig

In the right pane, right-click MaxPosPhaseCorrection, then Modify, in the Edit DWORD Value box, under Base, click Decimal, under Value Data, type a time in seconds such as 3600 (an hour) then click OK.

Now go back and click: HKEY_LOCAL_MACHINESYSTEMCurrentControlSetServicesW32Timeconfig In the right pane, right-click MaxNegPhaseCorrection, then Modify.

In the Edit DWORD box under base, click Decimal, under value data type the time in seconds you want to poll such as 3600 (polls in one hour) Exit Registry Editor

Now, to restart windows time service, click Start, Run (or alternatively use the command prompt facility) and type:

net stop w32time && net start w32time. And that’s it your time server should be now up and running.

This article explains the origins and workings of atomic clocks and how they are used to synchronise computer networks all over the world using NTP servers.

In conventional electronic clocks time is kept by running an electrical current through an oscillator which produces a repetitive electrical signal this is then governed by a quartz crystal to keep precision. These crystal oscillators are far more accurate than mechanical clocks but will still drift, perhaps over a second a week.

For day-to-day use crystal oscillators are a fine way to keep track of time; in the everyday running of our lives, a second makes very little difference, however, as light or radio waves can travel 300,000 miles in a second, some high technologies such as satellite navigation or global communication, require far more accuracy to be possible.

Atomic clocks are a timekeeping device that uses the known atomic resonance frequency of an atom to keep time. The first truly accurate atomic clock was built in 1955 at the National Physical Laboratory in the UK and was based on the caesium atom -133 which oscillates at exactly 9,192,631,770 every second.

This oscillation is actually a repetitive signal from the microwave radiation emitted by electrons in an atom when they change energy levels. Much of an atomic clock is designed to create the correct state to cause and augment oscillations.
Although other atoms can be used, the oscillation (9,192,631,770 a second) of the caesium -133 atom is now accepted by the International System of Units as being the definition of one second.

Atomic clocks are generally very large and constitute many highly technical apparatus such as vacuums and require whole teams of scientists to maintain and monitor the clocks. Much of which goes into compensating for unwanted side-effects such as frequencies of other atoms in the clock and even gravitational dilation (where according to Einstein’s theory clocks at different heights run differently because of the differences in the gravitational field) This makes atomic clocks highly expensive.

Fortunately many large scale national physical laboratories transmit radio time signals from their atomic clocks which can be used to synchronise standard crystal oscillators too.

Atomic clocks are also the basis of GPS (Global Positioning System) as each satellite contains an atomic clock as accurate time is integral for positioning (a position anywhere is made up of a direction, a velocity and time).
GPS signals can also be used to capture a time signal. This is now the most common way computer networks retain accurate time which is also essential in many communications and applications.

Most computer networks use a NTP server (Network Time Protocol) to synchonise their devices to an atomic time signal received via the GPS network.

A universal timescale, UTC (Coordinated universal Time), has been developed based on the time told by atomic clocks, TAI (International Atomic Time). UTC accounts for the slowing of the Earths rotation by adding leap seconds to TAI so as to prevent the gradual drift of night into day (although that would take 40,000 years or so) and allows the whole world to communicate using the same timescale.

This article explores the concept of keeping time and how human measurement of time is at odds with that of a computer.

Time is certainly a concept most of us take for granted, it passes us by and we only notice it when we catch a glimpse of a grey hair in the mirror or arrive late for that important meeting. Yet keeping track of the time has occupied mankind for millennia.
From early sundials and water timers to modern digital watches and atomic clocks, humans have found more and more accurate and innovative ways of telling the time.

Computers also need to know the correct time. Accuracy is essential in keeping the Internet and computer networks communicating with each other but to a computer the passing of time is a simple equation based on the accumulation of discrete moments added to a base time, normally the number of seconds from that point in time.

Humans on the other hand have a variety of different notions about how to measure time. We separate it in to seconds, minutes, days, weeks, months, years, decades centuries and even millennia.

And this is wehere the problem lies as historically we have forced time to correspond with the orbit and rotation of the Earth, called solar time, which as it turns out is not that precise, well not enough for a computer anyway.

Computer networks use Network Time Protocol (NTP), the time synchronization standard used by on the Internet to keep at the same time. NTP lets machines query regional time servers that get the Universal Coordinated Time UTC from highly accurate reference clocks either from the Internet or through radio or GPS receiver.

However, UTC is based on atomic time and it differs from the Earth’s rotational time (solar system) because the day is slowly lengthening. The moon’s gravity lengthens the global turn by roughly 1.4 milliseconds — that is, thousandths of a second — per day per century. Since 1820, what we think of as a 24- hour period has gotten 2 milliseconds longer.

As a result, atomic time differs from solar time by one second about every 500 days. To adjust leap seconds are added every year or so. However as computers become more reliant on accuracy this leap second can cause problems as a second can be a vasrt amount in some time sensitive applications.

Some suggest to combat this problem leap seconds should be eliminated and the world should stick with just atomic time even though that would result in sun at midnight and dark during the day (albeit in 43,000 years time). Others argue that having a time scale based on the Earth’s rotation is primitive and not needed in the modern age, although many farmers and astronomers are keen to argue the opposite.

However, as atomic clocks and computers become increasingly more accurate and precise it seems that humans and our spinning world are not going to be able to keep up.

Technology and the Importance of Time

This article explores the concept of time, how it is measured and how our technologies have required more and more accurate ways of measuring time.

It is a question that has perplexed philosophers and scientists since the dawn of man, ‘what exactly is time?’ and it has only been in our recent history that we have started to discover answers, thanks to Einstein and his work on special and general relativity.

We now know time is not the abstract concept we first thought it was, we also know it is not constant and is relative to different observers throughout the universe with the speed of light being the only constant in the universe.

In other words if the speed of light has to be the same for everybody then someone travelling at close to such a speed would find time slow down.

Fortunately as all humans live within the boundaries of the planet Earth it means the passing of time is very similar for us all (or so minutely different as to be impossible to measure). However, technologies such as satellites and GPS systems have to take into account this altering state of time otherwise they would become wholly inacurate.

As humans have progressed, telling the time with ever increasing accuracy has become more and more important. Historically, knowing the time was not so imperative. People needed to know the correct day to plant crops or when sunrise and sunset happened but accuracy was not a preoccupation.

However, since the invention of the mechanical clock followed at the turn of the twentieth century by electronic clocks, humans have started to rely on more and more accuracy for their technologies.

Seafaring, aviation and now space travel mean that humans have sought more and more accuarte ways of keeping time.

In the 1950’s atomic clocks were developed which were so accurate it was discovered that the revolution of the Earth, something we had based our timescale on for centuries, was no where near as accurate as these new clocks.

Now technologies such as the Internet, the Global Positioning System and satellite communication requires absolute precision as light can travel 300,000 km every second meaning accuracies of a split second could mean our satellite navigation systems could be out by thousands of miles and computer trading would be nigh on impossible.

Fortunately a global time scale, UTC (Coordinated Universal Time), has been developed and is based on the time told by atomic clocks. This allows systems all over the world to be synchronised to the exact same time.

Computer networks use the NTP protocol (Network Time Protocol) to receive a UTC timing reference and synchronise all machines on a network to that time.

NTP servers can receive a time reference over the Internet (although not very secure) from a national radio transmission (as long as the receiver is within range of a suitable transmission) or from the GPS network (via a rooftop GPS antenna).

Atomic clocks are incredibly expensive and generally they are normally only to be found in large scale physics laboratories such as MIT (Massachusetts Institute of Technology), NIST (National Institute of Standards and Technology (Colorado) or the National Physical Laboratory in the UK.

Fortunately many national laboratories broadcast the UTC (Coordinated Universal Time) time from their atomic clocks via a radio broadcast.

In the UK the national timing broadcast is called MSF and is broadcast by NPL (National Physical Laboratory) in Cumbria. The MSF broadcast is used by throughout the UK and parts of Europe to synchronise consumer electronic products like wall clocks, clock radios, and wristwatches. In addition, MSF is used for high-level applications such as network time synchronisation utilising NTP.

The time code contains the year, day of year, hour, minute, second, and flags that indicate the status of Daylight Saving Time, leap years, and leap seconds.

MSF operates on a frequency of 60 kHz and carries a time and date code that can be received and decoded by a wide range of readily available radio-controlled clocks and provides a received accuracy should be less than 10 milliseconds (1/100 of a second).

While many NTP servers now use GPS to receive a timing reference, the advantage of using a radio transmission is that a signal can be received indoors (a GPS antenna needs a good view of the sky).

However, the radio signal has a finite range and can be blocked by skyscrapers, mountains and dense conurbations. A radio based NTP server usually consists of a rack-mountable time server, and an antenna, consisting of a ferrite bar inside a plastic enclosure, which receives the radio time and frequency broadcast. The antenna should always be mounted horizontally at a right angle toward the transmission for optimum signal strength.

Similar national timing transmissions are broadcast from other countries in the US the signal is referred to as WWVB and is broadcast by the NIST (National Institute for Standards and Technology) in Fort Collins, Colorado, other systems are broadcast in Frankfurt, Germany (DCF-77), Japan (JJY) and France (TDF).

Atomic clocks are incredibly expensive and generally they are normally only to be found in large scale physics laboratories such as MIT (Massachusetts Institute of Technology), NIST (National Institute of Standards and Technology (Colorado) or the National Physical Laboratory in the UK.

Fortunately many national laboratories broadcast the UTC (Coordinated Universal Time) time from their atomic clocks via a radio transmission.

In the US the national timing broadcast is called WWVB and is broadcast by NIST (National Institute fro Standards and Time) in Fort Collins, Colorado. The WWVB broadcast is used by millions of people throughout North America to synchronize consumer electronic products like wall clocks, clock radios, and wristwatches. In addition, WWVB is used for high-level applications such as network time synchronization utilizing NTP.

The time code contains the year, day of year, hour, minute, second, and flags that indicate the status of Daylight Saving Time, leap years, and leap seconds.

WWVB broadcasts on 2.5, 5, 10, 15, and 20 MHz and for most users in the United States, the received accuracy should be less than 10 milliseconds (1/100 of a second).

While many NTP servers now use GPS to receive a timing reference, the advantage of using a radio transmission is that a signal can be received indoors (a GPS antenna needs a good view of the sky).

However, the radio signal has a finite range and can be blocked by skyscrapers, mountains and dense conurbations. A radio based NTP server usually consists of a rack-mountable time server, and an antenna, consisting of a ferrite bar inside a plastic enclosure, which receives the radio time and frequency broadcast. The antenna should always be mounted horizontally at a right angle toward the transmission for optimum signal strength.

Similar national timing transmissions are broadcast from other countries in the UK the signal is referred to as MSF and is broadcast by the National Physical Laboratory in Cumbria, other systems are broadcast in Frankfurt, Germany (DCF-77), Japan (JJY) and France (TDF)

Atomic clocks are incredibly expensive and generally they are normally only to be found in large scale physics laboratories such as MIT (Massachusetts Institute of Technology), NIST (National Institute of Standards and Technology (Colorado) or the National Physical Laboratory in the UK.

Fortunately many national laboratories broadcast the UTC (Coordinated Universal Time) time from their atomic clocks via a radio broadcast.

In the UK the national timing broadcast is called MSF and is broadcast by NPL (National Physical Laboratory) in Cumbria. The MSF broadcast is used by throughout the UK and parts of Europe to synchronise consumer electronic products like wall clocks, clock radios, and wristwatches. In addition, MSF is used for high-level applications such as network time synchronisation utilising NTP.

The time code contains the year, day of year, hour, minute, second, and flags that indicate the status of Daylight Saving Time, leap years, and leap seconds.

MSF operates on a frequency of 60 kHz and carries a time and date code that can be received and decoded by a wide range of readily available radio-controlled clocks and provides a received accuracy should be less than 10 milliseconds (1/100 of a second).

While many NTP servers now use GPS to receive a timing reference, the advantage of using a radio transmission is that a signal can be received indoors (a GPS antenna needs a good view of the sky).

However, the radio signal has a finite range and can be blocked by skyscrapers, mountains and dense conurbations. A radio based NTP server usually consists of a rack-mountable time server, and an antenna, consisting of a ferrite bar inside a plastic enclosure, which receives the radio time and frequency broadcast. The antenna should always be mounted horizontally at a right angle toward the transmission for optimum signal strength.

Similar national timing transmissions are broadcast from other countries in the US the signal is referred to as WWVB and is broadcast by the NIST (National Institute for Standards and Technology) in Fort Collins, Colorado, other systems are broadcast in Frankfurt, Germany (DCF-77), Japan (JJY) and France (TDF).

A GPS time server is really a communication device. Its purpose is to receive a timing signal and then distribute it amongst all devices on a network. Time server s are often called different things from network time server, GPS time server, radio time server and NTP server.

Most time servers use the protocol NTP (Network Time Protocol). NTP is one of the Internet’s oldest protocols and is used by the majority of machines that use a time server. NTP is often installed, in a basic form, in most operating systems.

A GPS time server, as the names suggests, receives a timing signal from the GPS network. GPS satellites are really nothing more than orbiting clocks. Onboard each GPS satellite is an atomic clock. The ultra-precise time from this clock is what is transmitted from the satellite (along with the satellite’s position).

A satellite navigation system works by receiving the time signal from three or more satellites and by working out the position of the satellites and how long the signals took to arrive, it can triangulate a position.

A GPS time server needs even less information and only one satellite is required in order to receive a timing reference. A GPS time server’s antenna will receive a timing signal from one of the 33 orbiting satellites via line of sight, so the best place to fix the antenna is the roof.

Most dedicated GPS NTP time servers require a good 48 hours to locate and get a steady fix on a satellite but once they have it is rare for communication to be lost.

The time relayed by GPS satellites is known as GPS time and although it differs to the official global timescale UTC (Coordinated Universal Time) as they are both based on atomic time (TAI) GPS time is easily converted by NTP.

A GPS time server is often referred to as a stratum 1 NTP device, a stratum 2 device is a machine that receives the time from the GPS time server. Stratum 2 and stratum 3 devices can also be used as a time servers and in this way a single GPS time server can operate as a timing source for an unlimited amount of computers and devices as long as the hierarchy of NTP is followed.

Telling the time is not as straight forward as most people think. In fact the very question, ‘what is the time?’ is a question that even modern science can fail to answer. Time, according to Einstein, is relative; it’s passing changes for different observers, affected by such things as speed and gravity.

Even when we all live on the same planet and experience the passing of time in a similar way, telling the time can be increasingly difficult. Our original method of using the Earth’s rotation has since been discovered to be inaccurate as the Moon’s gravity causes some days to be longer than 24 hours and a few to be shorter. In fact when the early dinosaurs were roaming the Earth a day was only 22 hours long!

Whilst mechanical and electronic clocks have provided us with some degree accuracy, our modern technologies have required far more accurate time measurements. GPS, Internet trading and air traffic control are just three industries were split second timing is incredibly important.

So how do we keep track of time? Using the Earth’s rotation has proven unreliable whilst electrical oscillators (quartz clocks) and mechanical clocks are only accurate to a second or two per day. Unfortunately for many of our technologies a second inaccuracy can be far too long. In satellite navigation, light can travel 300,000 km in just over a second, making the average sat-nav unit useless if there was one second of inaccuracy.

The solution to finding an accurate method of measuring time has been to examine the very small – quantum mechanics. Quantum mechanics is the study of the atom and its properties and how they interact. It was discovered that electrons, the tiny particles that orbit atoms changed the path that they orbit and released a precise amount of energy when they do so.

In the case of the caesium atom this occurs nearly nine billion times a second and this number never alters and so can be used as an ultra reliable method of keeping track of time. Caesium atoms are use din atomic clocks and in fact the second is now defined as just over 9 billion cycles of radiation of the caesium atom.

Atomic clocks are the foundation for many of our technologies. The entire global economy relies on them with the time relayed by NTP time servers on computer networks or beamed down by GPS satellites; ensuring the entire world keeps the same, accurate and stable time.

An official global timescale, Coordinated Universal Time (UTC) has been developed thanks to atomic clocks allowing the whole world to run the same time to within a few thousandths of a second from each other.

Methods of keeping track of time have altered throughout history with ever increasing accuracy has being the catalyst for change.

Most methods of timekeeping have traditionally been based on the movement of the Earth around the Sun. For millennia, a day has been divided into 24 equal parts that have become known as hours. Basing our timescales on the rotation of the Earth has been adequate for most of our historical needs, however as technology advances, the need for an ever increasingly accurate timescale has been evident.

The problem with the traditional methods became apparent when the first truly accurate timepieces – the atomic clock was developed in the 1950’s. Because these timepieces was based on the frequency of atoms and were accurate to within a second every million years it was soon discovered that our day, that we had always presumed as being precisely 24 hours, altered from day to day.

The affects of the Moon’s gravity on our oceans causes the Earth to slow and speed up during its rotation – some days are longer than 24 hours whilst others are shorter. Whilst this minute differences in the length of a day have made little difference to our daily lives it this inaccuracy has implications for many of our modern technologies such as satellite communication and global positioning.

A timescale has been developed to deal with the inaccuracies in the Earth’s spin – Coordinated Universal Time (UTC). It is based on the traditional 24-hour Earth rotation known as Greenwich Meantime (GMT) but accounts for the inaccuracies in the earth’s spin by having so-called ‘Leap Seconds’ added (or subtracted).

As UTC is based on the time told by atomic clocks it is incredibly accurate and therefore has been adopted as the World’s civilian timescale and is used by business and commerce all over the globe.

Most computer networks can be synchronised to UTC by using a dedicated NTP time server.