Why we need to increase the accuracy of our clock-watching
Image credit: Dreamstime
Timekeeping technology is struggling to keep up with the many aspects of modern life that rely on it, but more reliable and resilient solutions are on the horizon.
Even with the use of ultra-precise atomic clocks, time is a more fragile and unreliable commodity than we think. The UK government estimates that the time stamps used in London's financial trading are affected by between 80 and 120 GPS jamming incidents every month. In 2016, the decommissioning of a single GPS satellite led to 12 hours of IT and phone system errors globally.
The more efficient, integrated and faster our systems become - for communications, energy, transport, and financial services - the more dependent they are on these ultra-precise timings. With any slip in synchronisation, errors accumulate, resulting in disorder, damaging delays, unreliable records and a breakdown in trust.
Systems are fragile because all physical-measurement machines are prone to error when it comes to such high levels of precision. As global data signals have the potential to be jammed and spoofed by criminals and political enemies, access to atomic clocks can become dependent on stable political relations.
Take one of the most typical time sources, your computer. The system clock tick rate is just a microscopic tuning fork a couple of millimetres long. No two are identical - each will be affected by temperature in different ways and age differently. Typical drift is up to two seconds per day, meaning time stamps are wrong and interactions across networks are based on inaccurate timings. The Windows operating system refers to the Windows Time Service, based on a pool of Network Time Protocol servers, which again has been found to involve clock inaccuracies of more than a second.
The work of astronomers who calculated time from the movement of the Earth was replaced by International Atomic Time (TAI), using ultra-precise observations of the caesium-133 atom and 340 atomic time clocks in different locations internationally. The most relevant of the atomic clocks to our everyday lives are contained within Global Positioning System satellites. Information determining time to within 100 billionths of a second is shared globally wherever a signal can be received via satellites - making it unnecessary for individual organisations to operate their own atomic clocks.
Unfortunately, GPS can be cheaply and easily disrupted by transmitting noise on the same radio-frequency, while the satellites it relies on can malfunction or suffer interference from solar flares. However, it is relied on across public and commercial infrastructures. Mobile phones and other wireless comms depend on rigid time standards (an accuracy level of less than one second difference over 3,000 years) to put signals into order, prevent congestion and ensure calls and messages from different operators can be synchronised. The same applies to electricity power grids that have to link energy sources on the same frequency.
Accurate timestamps are critical for operations to be put into an order, to identify and avoid data bottlenecks, and to be relied on as evidence. Financial institutions and other businesses need to have a guarantee of the time and date of transactions to the microsecond and the order in which they took place: an explicit chain of cause and effect that establishes responsibility. Attempts to unpick the causes of the trillion-dollar ‘flash crash’ in the Dow Jones index in May 2010 were severely undermined by clocks that disagreed.
The increasing and intensifying level of demand for accurate timings has exposed the essential need for multiple solutions for when things go wrong: a failsafe backup system to correct any misalignments. One potential solution we are working on at Cranfield University seeks to establish technology that can underpin trusted and reliable navigation, communication and surveillance, as well as timings for traffic management in digital aviation and transport in smart cities.
The work builds on existing tech created by UK company Hoptroff and its experience developing Cloud-based backup for the financial services sector. The system is based on a network of mutually resilient cloud-timing hubs, each consisting of three nanosecond-accurate ‘grandmaster’ clocks connected to three different sources. The hubs continuously compare the different sources to ensure that accuracy and source traceability are maintained. To mitigate against satellite communication issues, a backup is supported by a terrestrial location at RISE, the Research Institutes of Sweden.
The accuracy of GPS is not being monitored to the extent needed for resilience. That’s dangerous. We need to take on the challenge of how national and linked international infrastructures can always depend on time data and that means recognising the problem and investing in new ideas - in particular, being open and alert to emerging innovations coming through from small and micro-enterprises.
Dr Simon Harwood is director of defence and security at Cranfield University, where Dr Ivan Petrunin is a lecturer
Sign up to the E&T News e-mail to get great stories like this delivered to your inbox every day.