Synchronization plays a critically important role in telecom networks. Mobile networks rely on accurate synchronization to perform functions such as handovers between cell sites, to minimize interference that stems from joint transmission and reception across sites, and to maximize performance at the cell edge. Synchronization ensures that the frequency, phase and time differences among systems remain within tolerable limits to maintain proper network operation.

As network operators move to 5G networks and make greater use of TDD spectrum, the need for accurate phase and time synchronization is growing. For example, TDD radios use the same frequency for uplink and downlink transmission, so they require time and phase synchronization in addition to frequency synchronization.

While synchronization can be achieved by placing GNSS receivers at each cell site, there are cases where this approach is not viable because of GNSS receiver costs, line-of-sight limitations, or a desire to mitigate jamming or spoofing. As a result, network operators need an alternate backup or primary network-based synchronization source.

Many operators are turning to network-based synchronization distribution using packet transport networks that support a combination of the Precision Time Protocol (PTP) and enhanced Ethernet equipment clocks (eEECs). This hybrid model enables the network to deliver highly accurate synchronization that meets the requirements of all time-sensitive applications.

The need for higher synchronization accuracy

In 4G and 5G mobile communications networks, some applications require coordination among the RAN elements. Absolute or relative synchronization limits apply to specific applications as summarized in Table 1.

Table 1: Frequency and phase/time sync requirements in telecom networks

For frequency synchronization, the air interface requires an accuracy of 50 ppb regardless of the application. However, the accuracy requirements for phase/time synchronization vary depending on the RAN application being used. For example, radios that operate in TDD spectrum have a general requirement to be within +/- 1.5 msec absolute time error. Applications that use coordinated RAN features such as intra-band contiguous carrier aggregation (CA) require much stricter synchronization accuracy within +/- 130 ns relative time error (between the coordinated cells). Failure to meet these tolerances can result in poor signal quality, lower data speeds, interference and network outages.

Other applications impose tight accuracy requirements on the nodes that participate in the synchronization distribution, where each additional nanosecond of time error impacts performance. These include teleprotection applications within power utility networks, where a highly accurate time reference is useful in fault analysis because the location of faults can be found by analyzing timestamped measurement values such as precise voltage measurements.

Enhanced 911 applications also have tight requirements. Government mandates in the US call for positioning accuracy within 50 meters for emergency calls that originate from wireless devices. To achieve this level of accuracy, the total time error must be minimized as the position of the wireless device will deviate approximately 0.3 meters for every nanosecond of time error. Therefore, techniques that improve the timing accuracy of the nodes that participate in the synchronization chain can lead to improved performance.

The role of eEEC in enhancing phase/time synchronization accuracy

Telecom networks use a hybrid model in which Synchronous Ethernet (SyncE) assists the PTP in generating more accurate timing. SyncE adds physical-layer clock distribution and synchronization to Ethernet. This physical layer frequency is provided to the PTP time clock, where it serves as the stable reference frequency. The reference frequency can be traced to the primary reference clock (PRC). It is critical for maintaining PTP accuracy.

In SyncE networks, slave clocks are locked to the frequency of the PRC, where they recover the clock from the incoming physical layer signal. However, high-frequency variations (jitter) and low-frequency variations (wander) caused by network device processing, temperature changes and other factors can impact the timing signal quality. Therefore, jitter and wander should be minimized in network devices that process clock signals to avoid mistiming the transmitted signal and hindering the distribution of the clock reference across the network.

In support of packet-based network synchronization, the boundary clock and transparent clock functions are defined within the IEEE 1588v2 PTP protocol. Use of these clocks avoids the accumulation of packet delay variations (PDVs) that would adversely impact the PTP messages as they transit non-PTP routers and switches.

The PTP protocol is designed to interwork with existing frequency synchronization mechanisms such as EEC or eEEC, where the PTP-aware nodes operate in a “SyncE assist” mode to improve PTP performance. To meet the stringent timing requirements of fronthaul and other applications, packet transport networks must use Class C Telecom Boundary Clocks (T-BCs), as recommended by the IEEE 802.1CM/CMde TSN for fronthaul specification. Class C clocks use eEEC instead of EEC to increase synchronization accuracy.

Benefits of eEEC

In packet-based networks, eEEC delivers much better performance than EEC. The main advantages include:

• Better frequency synchronization performance: eEEC has stricter clock frequency performance requirements for jitter and wander and improves clock stability. It provides a roughly five-fold increase in performance for wander generation compared to EEC and increases synchronization accuracy by reducing the time error.
• Better phase/time synchronization performance: SyncE helps minimize the phase/time error for PTP to reach Class C T-BC. It provides a physical layer source that enables frequency alignment to improve phase/time synchronization performance.
• Improved time noise generation, transfer and holdover of T-BCs: SyncE improves the time noise generation, transfer and time holdover of SyncE-assisted T-BCs. SyncE with eEEC enables an improved noise transfer bandwidth of 1–3 Hz versus 1–10 Hz for EEC. In addition, SyncE with eEEC improves the holdover for an enhanced equipment clock by reducing the permissible phase error.
• Longer chain or performance that is better than +/-1.5 µs performance: Because of the increased synchronization accuracy possible with enhanced SyncE-assisted T-BC clocks, each Class C T-BC node conforms to a +/-10 ns cTE and 30 ns max|TE| time error. This enables longer chains and more robust architectures with support for transport path diversity.

Figure 2: Performance improvements in moving from EEC to eEEC

Meeting the synchronization needs of 5G networks

High-quality network synchronization is of paramount importance to the proper operation of 5G networks, particularly for time-sensitive applications, architectures and features that impose strict synchronization accuracy requirements.

A hybrid model that uses SyncE physical layer frequency to assist PTP helps minimize the phase/time error for PTP to reach Class C T-BC or better. This approach enables the packet transport network to meet or exceed the stringent synchronization demands of time-sensitive applications such as those in the fronthaul segment.

Find out more

White paper: Delivering high-quality network-based synchronization