The road to 5G is paved with flexible PTP profiles

Migrating telecommunications, utilities, transportation, the defense sector and other critical infrastructure to 5G has spurred new phase requirements and implementation of the Open RAN system to support interoperability between cellular network equipment provided by different vendors. As a result, the industry has standardized on not just one but two precision time protocol (PTP) profiles and requires a synchronization architecture that can handle increased mobile network complexity.

There is no one-size-fits-all solution to these challenges of adding 5G mobile services to 4G networks. However, operators can often leverage and successfully build upon existing synchronization investments.

The synchronization challenge

5G mobile synchronization has become more granular, and it is no longer possible for operators to universally adopt the ITU-T G.8275.1 PTP profile. Instead, 5G technology’s use of time-division duplex (TDD) brings both relative and absolute new phase requirements, and Open RAN introduces baseband unit (BBU) functions that are disaggregated into radio units (RUs), distributed units (DUs) and centralized units (CUs). As a result, the telecom industry has standardized two PTP profiles to address PTP-aware and PTP-unaware networks, respectively: ITU-T G.8275.1 and ITU-T G.8275.2.

To navigate this transition, it is important to understand how the existing 4G synchronization architecture operates.

Building on a 4G foundation

Starting with the backhaul network for 4G services, operators know that packet-delay variation can have a major impact on synchronization. In many countries, PTP was implemented as a 4G backup synchronization mechanism using a global navigation satellite system (GNSS) as the primary synchronization source. To maintain phase services in the event of a GNSS failure, a PTP flow was adopted by the ITU-T as G.8273.4 (Assisted Partial Timing Support, or APTS) to connect the edge primary reference time clock (PRTC) to the centralized core clock.

Figure 1: A typical mobile 4G synchronization architecture employs a grandmaster clock and G.8275.2 PTP profile. (Source: Microchip Technology Inc.)

In this architecture, the PTP flow is essentially a proxy GNSS signal from the core that has traceability to universal coordinated time (UTC). The PTP input is calibrated for time error using the local-edge PRTC GNSS, which has the same UTC reference as the upstream GNSS.

While a typical 5G architecture has a number of key elements, as shown in Figure 2, it is essential that operators consider these backhaul dynamics, as well as those in the fronthaul and mid-haul networks that 5G disaggregation introduces.

Figure 2: The disaggregated BBU in an Open RAN 5G network architecture adds network devices and increases synchronization complexity. (Source: Microchip Technology Inc.)

The fronthaul becomes the network synchronization focal point for serving 5G RUs or 5G base stations. This fronthaul network uses the G.8275.1 multicast profile to serve 5G base stations (gNodeBs), and PTP provides the primary synchronization mechanism. Operators need to consider the end-to-end timing budget, which should be ±1.5 µs, as well as the relative-time accuracy between adjacent RUs, which should be 130 ns/260 ns.

In contrast with this G.8275.1 profile, the ITU-T G.8275.2 profile exists at Layer 3, unicast. There is no need to provide on-path support capability on all network elements because the PTP protocol flows through them as high-priority traffic. The network’s primary need in this use case is adequate PTP client capacity support from the PTP grandmaster. A good capacity number is typically 100 clients, minimum, and up to several thousand in some cases.

There are a few other fronthaul aspects to understand from a synchronization standpoint. First, synchronization operates from a source of time from a GNSS signal, with APTS protection when the GNSS signal is either unavailable or intermittent. Second, fronthaul is generally located in large cities and metro areas that contain many base stations served by nearby PTP grandmasters. The network uses a G.8275.1-based, telecom-specific PTP profile, and network elements operate in multicasting mode to embed a modern boundary clock. This approach does not require significant capacity.

Outside of metro areas, PTP provides frequency synchronization using grandmaster clocks to serve what are primarily older FDD radio systems. These clocks coexist with a mix of older radios and new environments that have been deployed as part of moving to 5G.

To meet 5G’s timing and phase accuracy requirements, many operators are migrating their frequency-focused grandmasters to newer generations of IEEE 1588 PTP grandmasters. These clocks expand overall capabilities as well as the number of PTP ports compared with prior solutions and must connect to many more devices, such as older radios and cell towers or other PTP grandmasters.

Backhaul sites and grandmasters typically use the ITU-T G.8275.2 profile, which runs at the Internet Protocol (IP) layer. While the telecom industry focuses on enabling migrations of legacy environments toward newer architectures and devices, they must still support legacy signal systems, such as synchronization supply units (SSUs) and primary reference clocks (PRCs). These systems will need to be integrated into the newer 5G- and PTP-focused architectures.

One other consideration is the ability to integrate systems that are located at a great distance from the grandmasters.

Making the transition while preserving investments

Typically, large operators who add 5G mobile services to their 4G networks will install PTP grandmasters in their central offices. This enables them to support wireline broadband and wireless mobility across four typical use cases.

The first use case is replacing the legacy PRS systems and migrating to a newer-generation grandmaster that can function as a PRS or enhanced PRS (ePRS). This is a typical use case in North America. The second use case is to migrate directly from a traditional PRTC grandmaster to a more modern platform. This increases connectivity options while delivering advanced APTS capabilities. It also provides frequency synchronization for cell-site backhaul across thousands of clients using PTP G.8275.2.

A third scenario is deploying new PRTC grandmasters for 5G fronthaul using PTP G.8275.1. Alternatively, a fourth scenario is to migrate existing synchronization systems to more modern and resilient PTP grandmasters. This enables operators to meet stringent 30-ns accuracy to UTC while achieving 14 days of holdover where needed.

Each of these installation options preserves investments while enabling operators to leverage newer technologies over time. They provide a path to serving 5G sites by evolving their existing synchronization infrastructure.

Other considerations

Aside from fronthaul and backhaul considerations, there are several other issues to address when choosing a timing profile and capacity requirements.

For instance, some countries or operators may not own the infrastructure used for some or all of their deployments. This is often the situation in North America, where operators commonly lease their backhaul lines from third parties. It is not always the case that these leased lines meet the operator’s time and phase performance requirements. As a result, they may not be able to rely on the backhaul links. Additionally, they may not be able to monitor the leased lines’ synchronization quality.

To better serve these mobile operators, leased-line backhaul providers are upgrading their network elements with boundary clocks that deliver highly accurate time and phase. This ensures they can meet the stringent timing requirements for 5G architectures. New entrants, such as satellite providers or cable operators, have similar concerns as they add mobile to their portfolio over leased architecture.

As for legacy wireline providers that lease their wireline infrastructure to mobile operators and new mobile entrants, it may be necessary to upgrade their infrastructure to deliver more accurate time and phase so that they can run either G.8275.1 or G.8275.2 over the leased backhaul layer with a guaranteed level of time accuracy.

The right PTP profiles for the purpose

One PTP profile does not fit all needs, nor does one PTP capacity level. Mobile operators should examine all options as they implement a 5G architecture or launch a 5G service. These options are based on standards that can be deployed at both the fronthaul and backhaul networks.

As operators assess these 5G service migration options while also considering their specific regional, network transport and integration requirements, they will typically want to employ various PTP profiles and capacity levels. This can often be achieved with a migration strategy that preserves their existing synchronization investments.