Blueprint for network synchronization
Telecommunications networks are shifting from circuit- to packet-switched technologies to meet exploding demands for bandwidth, but transitioning to packet-based networks requires a change in synchronization architecture. Asynchronous Ethernet networks do not provide physical circuits between network elements, and consequently, base station sync must be engineered into the packet backhaul with a timing technology such as Precision Time Protocol (PTP) and a holdover backup.
As network operators look to design the sync architecture for 4G/LTE networks, considerations must be made to meet the stricter synchronization requirements necessary to support the latest mobile technologies in conjunction with Location-Based Services (LBSs) such as E911. This will require network designers to implement a synchronization solution that can support both frequency and phase. Engineers must ensure synchronization interoperability among network equipment manufacturers which will be key to successful deployment, and lastly, they must define and architect a sync backup in the event the primary sync signal is lost to ensure continuity of service.
The sync switch
Mobile base stations that rely purely on frequency control, such as GSM and UMTS, have frequency accuracy requirements of 16 parts per billion (ppb) for G.823-compliant physical layer clocks on the E1/T1 backhaul connection (the transport interface). This requirement ensures the clocks’ internal oscillators lock and generate the 50 ppb accuracy required to align the base stations with mobile phones at the RF layer (the air interface) and avoid dropped calls.
As the backhaul transitions to Ethernet for packet processing, the TDM physical layer synchronization service chain is broken. The loss of physical layer sync generates the necessity for base station designs to incorporate PTP slave clocks that meet the 16 ppb requirement using packet technology. In base stations, such slaves rely on access to a carrier-grade PTP grandmaster clock deployed in the Mobile Switching Center (MSC) or Radio Node Controller (RNC). Figure 1 shows a typical example of PTP synchronization for cellular networks.
With the network transitioning to 4G/LTE TDD (Time Division Duplex), more stringent phase synchronization is now required to support the tighter use of frequencies and emerging LBSs, including E911 requirements. IEEE 1588-2008 PTP is a next-generation packet-based timing protocol targeted for use in asynchronous network infrastructures based on packet transport technologies. PTP has gained traction as the technology of choice to deliver synchronization for packet-based networks because it delivers both the frequency and phase synchronization required for 4G/LTE networks and the transition to them.
Precise sync, and the Telecom Profile
For any technology to be successful, it must be standards-based and interoperable between equipment vendors. To meet this requirement, the IEEE 1588-2008 introduced the concept of profiles to specify combinations of options and attribute values to support a given application.
The proliferation of PTP sync technology in over 100 networks worldwide was due in large part to having a standards-based technology supported by the “Telecom Profile,” which defined the parameters for interoperability among vendors. The purpose of the profiles is threefold:
· Profiles were introduced in IEEE 1588-2008 to allow other standards bodies to tailor PTP to particular applications
· Profiles contain a defined combination of options and attribute values aimed at supporting a given application
· Profiles allow interoperability between equipment designed for a given application
The “Telecom Profile,” released by the ITU-T as recommendation G.8265.1 in October 2010, addresses the application of the PTP to the frequency synchronization of GSM, UMTS, and LTE-FDD base stations.
Backing up the sync signal
Depending on the type or geographic location, some networks have relied heavily on GPS technology to deliver synchronization. However, GPS synchronization is susceptible to jamming and spoofing or simple signal fades where antennas are partially blocked and network sync is disrupted.
Regardless of the primary technology used to synchronize the packet-based network (PTP or GPS), however, holdover technology performs a critical function within the specified requirements of base stations to support 4G/LTE services.
Holdover req’s and tech’s
Holdover is the period of time required to keep network sync stabilized when the sync source is disrupted or unavailable. Holdover is achieved by equipping cellular Base Transceiver Stations (BTSs) with oscillators or atomic clocks that temporarily “holdover” sync signals. Holdover periods can range from several hours to several days depending on the oscillator technology (crystal or rubidium), environmental factors (temperature and temperature variation), and the quality of the implementation (algorithms that account for and adapt to the effects of aging).
Holdover requirements vary depending on type, complexity, and operator requirements. 4G/LTE Time Division Duplex (TDD) networks have more stringent timing requirements than 2G/3G networks, and some applications – such as LBSs and E911 – impose even more exacting sync requirements to accurately locate handsets through triangulation from base stations.
It must be noted that different grades of oscillators deliver varying holdover performance, which of course will also vary in cost. To ensure continuous network operations, it is recommended that service providers deploy rubidium atomic clocks in their base stations to ensure holdover for either a GPS or a PTP synched network (Figure 2).
Rubidium is the best technology available on the market today and delivers holdover of up to 1.5 µs (required for LTE-TDD) over a 24-hour period. Even for 2G/3G environments where the accuracy requirements are not as stringent, rubidium provides a significant advantage, as the much longer holdover period can save weekend or nighttime truck rolls. Important factors favoring rubidium holdover technology are:
· Rubidium atomic clocks have a much faster lock time, thereby reducing the impact of a power outage or other event that would require systems to cycle.
· Innovation has yielded reduced size and lower power consumption, making rubidium solutions easier to embed in equipment designs.
· Costs for rubidium clocks are on a steep decline: five years ago prices were double what they were two years ago, and technical innovation continues that downward trend today.
The point is that under similar environmental circumstances and within the price/performance ranges targeted for base stations, rubidium provides holdover performance significantly better than other oscillators.
Sync and holdover for the future
Carrier service availability has always relied on redundancy and backup solutions to meet the expectations of their customers. A reliable end-to-end synchronization solution for a packet-based network requires the use of a primary sync source (such as IEEE-1588 PTP) and an embedded rubidium atomic clock at the base station. In this solution, multiple technologies aid one another to extend the holdover time of rubidium and allow installation of base stations in locations that were not practical in the past. This approach has been tested for deployment and is ready for carriers to include in 4G/LTE build-out plans.
With the evolution of mobile networks to 4G/LTE, the requirements for synchronization become more stringent with the advancement of tighter phase requirements. To meet the phase requirements of ±1.5 μs and ensure continuous network operations, rubidium atomic clocks are required to deliver network holdover as a backup to either PTP or GPS-based synchronization technologies.