Time synchronization in telecommunication networks

Now with the arrival of 5G the need for synchronization in the RAN has grown. Why? One reason is greater use of spectrum based on time-division duplexing (TDD). Another is that new radio technologies and network architectures have emerged to boost efficiency, as well as support demanding 5G use cases. Each of these developments throw the spotlight on synchronization.

 

Smart factories and industrial automation, for example, where robots, machines and IoT devices are ‘talking’ to each other, need time-sensitive networking (TSN). Autonomous vehicles is another important TSN use-case. If Industry 4.0 and 5G ‘network slicing’ are to realize their full potential and generate new revenue streams for mobile network operators, ops teams must first make sure their timing is not ‘off’.

 

Synchronization requirements between TSN nodes, as pointed out by Helmut Imlau, a timing and synchronization expert at Deutsche Telekom, must be measured in the “hundreds of picoseconds” if 5G is to ensure “full performance”.

 

Tomasz Widomski, an engineering chief expert at ELPROMA, a specialist in network synchronization, goes as far as to recommend the immediate suspension of the so-called “leap second”, a one‑second adjustment occasionally applied to Coordinated Universal Time (UTC). The adjustment is to accommodate the difference between “precise time” as measured by atomic clocks – ELPROMA’s favored timing mechanism –and imprecise observed “solar time”.

 

In a letter to the ITU, Widomski warned that continued use of the leap-second “introduces a high risk of failure for IT and Industry 4.0”.

Telcos need time on their side

 

Synchronization in telco networks is the process of aligning time scales of transmission and switching equipment. When this is done to high levels of accuracy, operators can have peace of mind that operations happen in the correct time and — just as importantly — in the right order. There is harmonization.

 

As Ericsson recently pointed out, it is not that fundamental synchronization requirements have become more stringent it is rather they have become more critical with the arrival of new technologies and use cases.

 

Take autonomous networks. In 2019, the TM Forum, in collaboration with BT, China Mobile, Huawei, Orange and Telstra, published a whitepaper outlining their vision of what autonomous networks — where the need for manual human intervention is drastically reduced — might look like.

The whitepaper’s authors want network planning, design and configuration to be automatically performed, and service rollout time shortened by as much as one tenth. Moreover, using so-called “closed-loop mechanisms” — which essentially means the network looks after itself in terms of service fulfilment and assurance  — the group set out a firm industry direction towards achieving “optimal user experience”. This alluring vision will not be realized, however, without reliable and accurate time synchronisation between different network elements across large areas.

 

The same is true with 5G. RANs, both for 4G and the next-gen tech, are designed to optimize service performance and reliability, and synchronization is a vital enabler. This is especially the case when supporting new and distributed RAN architectures, where the 5G RAN is separated into different logical units: the centralized unit, the distributed unit and the radio unit.  

 

Moreover, many commercial 5G networks use TDD radio frames, which inherently require time and phase alignment between radio base stations to prevent interferences and loss of traffic. Time synchronization is also needed in networks using frequency-division duplexing (FDD) when different radio coordination features are used.

 

The UK’s BT is one example of a network operator paying more attention to synchronization, which now relies on  “high-capacity synchronisation technology” to make better use of its spectrum assets for both 4G and 5G.

 

The new timing solution distributes stable and sub-microsecond accuracy phase and time-of-day information, and replaces BT’s old timing network based on frequency synchronisation. BT said it will use the tech to roll out 4G coverage in areas with little or poor coverage, and to eventually deploy 5G services across the UK. BT Group Chief Architect Neil McRae said the new timing approach was “key to our 5G aspirations”.

How do you want to clock-in?

Network operators have various timing protocols they can draw upon to meet synchronization requirements between network elements and ‘grandmaster clocks’ (grandmaster clocks receive UTC information from an external time reference, most commonly a global navigation satellite system (GNSS) source).

 

The Precision Time Protocol (PTP) developed by ‘IEEE 1588’ for local systems needing greater accuracy than can be provided by the Network Time Protocol (NTP) – NTP was developed by David Mills at the University of Delaware in 1981 to be fault-tolerant and scalable – is one protocol. Primary reference time clock (PRTC) requirements, set out by ITU recommendation G.8272, is another.

There is also an enhanced PRTC solution (ePRTC), which is designed to protect governments, network operators and enterprises from regional, and even global, GNSS outages.

 

By producing an independent time-scale, which is aligned with GNSS but simultaneously maintaining the autonomy of ePRTC-generated phase, time, and frequency signal outputs, the solution is designed to reduce GNSS dependency. This is something ELPROMA strongly advocates. ePRTC enables a network operator to take back control of the timing source used in the event of GNSS failure.

 

Given the importance of network synchronization, Suntech has teamed up with ELPROMA and proved that time synchronization equipment and connectivity can be included in the SunVizion Network Inventory Management System. Along with all other network assets, spread across different domains, SunVizion provides comprehensive visualization and up-to-date performance checks of synchronization equipment.