The Synchronization Problem in 5G Networks

Published in Telesemana by Pablo Bertrand

by Pablo Bertrand

Published in Telesemana

In next-generation mobile networks, precise time synchronization is not a technical luxury but a necessity. It has a direct impact on service quality, spectral efficiency, and the interoperability of multiple technologies within the fifth generation of 3GPP communications. For example, in 5G radio access architectures based on Time Division Duplex (TDD), all neighboring cells must switch between transmission and reception in perfect coordination to avoid mutual interference. A temporal deviation of just microseconds can cause overlaps in transmission cycles, generating packet losses, dropped calls, or a substantial degradation in data rates. This was not necessarily the case in LTE, as Frequency Division Duplex (FDD) was predominantly used. The bar is now higher.

Looking at the near future of the region, the use of technologies such as massive MIMO, beamforming, and ultra-dense networks with frequent handovers requires that base stations share precise information about events occurring in brief time intervals. In networks where resources are dynamically allocated and coordination algorithms are time-sensitive, the lack of synchronization can lead to inefficient spectrum allocation, increased latency, and a general deterioration of the user experience. At the service level, poor synchronization can prevent the operation of technologies such as carrier aggregation, which increases throughput by receiving from more than one base station. In this context, synchronization is a structural requirement for the 5G ecosystem to fulfill its promises.

At the core of this synchronization is the Precision Time Protocol (PTP), standardized as IEEE 1588. PTP operates with a master-slave architecture in which a Grandmaster clock provides a time reference to slave clocks distributed across the network. This protocol achieves synchronization precision on the order of nanoseconds, a standardized requirement for 5G applications that demand strict time margins. To address different network scenarios, specific PTP profiles have been developed, such as those defined by ITU-T recommendations G.8275.1 and G.8275.2.

When implementing PTP in 5G networks, operators generally consider two main approaches: the use of centralized clocks or the strategy known as "GNSS Everywhere."

"GNSS Everywhere" involves equipping each base station with its own GNSS receiver, providing direct access to precise time information. This approach simplifies the synchronization architecture and eliminates dependence on intermediate network elements, while also being initially more cost-effective in cases where base stations already have the technology. The major drawback of this approach is that GNSS signals are susceptible to interference and even spoofing attacks, raising concerns about reliability and security.

On the other hand, the centralized Grandmaster clock approach involves deploying high-precision clocks, usually synchronized via GNSS, at strategic locations within the network. These Grandmasters distribute time information across the network using PTP, utilizing transport network elements as boundary clocks to maintain precision. This option is more reliable, as the oscillators in Grandmaster clocks can use materials such as Rubidium or Cesium, which are capable of maintaining clock signal precision even after losing the GNSS signal, through a characteristic known as holdover (better oscillators will provide greater holdover). However, it requires intermediate network elements to be PTP-compatible, in addition to the oscillators themselves being geographically distributed, which can involve costly hardware upgrades.

In the context of the inherent limitations of current synchronization methods, the ITU has begun standardization to introduce a new approach called Enhanced Partial Timing Support (ePTS). This proposal aims to provide time synchronization without requiring full support on all intermediate nodes. ePTS operates as an overlay on the network to distribute timing information, reducing the need for hardware upgrades and maintaining time signal resilience against GNSS system vulnerabilities. By enabling synchronization over existing infrastructure, ePTS is emerging as a promising solution for operators seeking to prioritize cost and complexity reduction without compromising reliability in the 5G context.

However, this is not yet a finalized standard, so operators deploying 5G in Latin America who prioritize cost reduction face a dilemma: either depend on GNSS reception at each base station, since their vendors likely support it and they would not need to incur additional costs at the expense of a less reliable system, or invest in clocks with higher-quality oscillators.

Today, a software-defined radio device can generate a signal that interferes with GNSS and disable 5G mobile service without PTP backup for less than 300 dollars. Even without considering intentional attacks, the simple loss of a GPS satellite signal could result in the failure of multiple cells. This demonstrates the fragility of architectures that depend exclusively on GNSS for time synchronization.

In this scenario, it is essential that operators incorporate PTP-based backup mechanisms to guarantee the operational continuity of their services. Investing in a robust synchronization architecture, capable of fully or partially supporting time distribution, is a strategic investment to protect the reliability of the 5G ecosystem against critical and easily exploitable vulnerabilities.