5G synchronization is an essential factor of next-generation 5G networks since new 5G applications will emerge with extremely low latency requirements and very strict specifications. 5G NR is mainly a TDD based system that requires tight synchronization such as time and phase for proper operation and guaranteed network performance. Therefore, new standards and protocols such as 1588v2 Precision Time Protocol (PTP) with nanosecond accuracy are increasingly critical to the success of high-speed, low-latency 5G applications.
Synchronization basics
Traditional telecommunications networks were using time-division multiplexing (TDM) technologies such as T1/E1 and SONET/SDH. T1/E1 was based on PDH (Plesiochronous Digital Hierarchy) technology, which was “almost” but “not quite” synchronized. Synchronous Digital Hierarchy (SDH) or Synchronous Optical Networking (SONET) were optical backbone networks tightly synchronized by external atomic clocks.
Over the last decade, the flexibility and higher data rates of packet networks have already pushed Telecom Operators to migrate to IP Backhaul/Transport networks. IP networks have enabled the industry to support new advanced features and benefit from improved performance and network scalability. However, packet networks are by default, asynchronous, and suffer from packet delay variation, which poses a significant threat to overall network performance. This makes network synchronization a quite demanding aspect for every Mobile Operator.
Packet networks may be synchronized via Frequency, Phase, and Time. In frequency synchronization, the reference and measured pulses have the same frequency, which means that they are spaced in equal intervals but not at the same moment. In phase synchronization, the reference and the measured pulses are equally spaced and at the same moment. Finally, in time synchronization, the reference and the measured pulses have the same frequency and phase, which means they are equally spaced at the same moments, but also at the identical time.
Transport networks achieve the required precision and network accuracy by using an appropriate primary timing reference source. The primary source can be provided by atomic clocks such as Caesium, Rubidium, and Quartz crystal. Another common clock source is a signal provided by a Global Navigation Satellite System (GNSS). This system transmits geolocation and time information to a GPS (Global Positioning System) receiver that can later be used for device synchronization.
Synchronous Ethernet
Synchronous Ethernet (Synch-E) is a physical layer-based frequency synchronization technique that provides frequency synchronization to the packet layer of Ethernet-based networks. Synch-E provides a stable and accurate frequency reference, which is derived by a high-quality clock source. This frequency reference is then passed to the transport elements to synchronize the internal clock of each Ethernet node. Synch-E is used to synchronize only in frequency and needs support from all the elements in the transport chain.
Synchronous Ethernet is specified in various ITU standards, which define clock specifications, jitter and wander values, clock quality levels, etc. ITU G.8261 “Timing and Synchronization Aspects in Packet Networks” describes the architecture of Synch-E as an extension of existing SDH/SONET networks to transport synch messages via Ethernet building blocks. ITU G.8262 “Timing Characteristics of Synchronous Ethernet Equipment Slave Clock” specifies compatibility of Synch-E with SDH clocks and defines parameters such as clock accuracy, tolerance, etc. Finally, ITU G.8264 “Distribution of Timing through Packet Networks” extends the Synchronization Status Message (SSM) functionality to Synchronous Ethernet and specifies the Ethernet Synchronization Messaging Channel (ESMC).
Phase and Time Synchronization
Unlike Synch-E that supports frequency synchronization, a packet-based system may also synchronize in phase and time. The Precision Time Protocol (PTP) is specified in IEEE 1588 “Standard for a Precision Clock Synchronization Protocol for Networked Measurement and Control Systems,”. PTP is defined as the protocol to transfer Phase and Time of Day (ToD) information over packet networks and synchronize separate nodes.
The initial 1588 standard was published in 2002, officially entitled IEEE 1588-2002, while it was superseded by the 2008 version, also known as 1588v2. IEEE 1588 (also referred to as plain PTP) can be used by a system with very strict synchronization requirements. That is why it is the main candidate for next-generation communication networks such as 5G. PTP is the ideal alternative in networks where the use of a GPS in every node is either very costly or satellite signals are not accessible.
Grandmaster Clock
The main concept of 1588v2 is that the clock timing reference, such as a GPS or atomic clock, serves as the Grandmaster Clock. Timing reference is forwarded to the transport nodes via PTP exchange messages in a Master and Slave fashion.
Ordinary Clock
An Ordinary Clock (OC) is an element that has a single physical PTP port that is used to exchange synchronization messages. It usually serves as the end node within a synchronization network, and that is why in a mobile system, typically, all the slave clocks of the Mobile Base Stations are ordinary clocks.
Boundary Clock
A Boundary clock (BC) has several physical PTP ports and may terminate and generate PTP flows towards all slaves in the chain. As opposed to an ordinary clock, a boundary clock may regenerate the reference coming from a grandmaster. It may then distribute the synch messages to several slave ports, which makes it more suitable for a nodal or aggregation node.
Transparent Clock
A Transparent clock (TC) does not generate PTP messages but acts as a typical “transparent” bridge/router that handles and forwards PTP packets. End-to-end Transparent Clock registers the arrival time of a PTP message and uses the correction field to reduce packet delay variation (PDV). Peer-to-peer Transparent Clock computes the link delay between the local port and the port on the link peer, which is the equivalent port on another node of the link.
PTP network performance
Time Error in nanoseconds (ns) is the main measurement unit that captures the time difference between two clocks. Time error is generally measured using three basic parameters. Maximum Absolute Time Error (max|TE|) is the maximum distance of the time error value from zero, while Constant Time Error (cTE) is the mean value of the time error function, both measured in nanoseconds. Dynamic Time Error (dTE) is the change of the time error function, which actually represents the phase or time wander and is specified with MTIE & TDEV masks.
According to Recommendation ITU-T G.8273.2/Y.1368.2 “Timing characteristics of telecom boundary clocks and telecom time slave clocks” in order to support different performance requirements at the end application, the maximum absolute time error, the time error and dTE noise generation requirements for T-BCs are divided into two classes: Class A and Class B. At the precision time protocol (PTP) and 1 pulse per second (PPS) outputs, the maximum absolute time error (max|TE|) for Class A is 100ns and Class B 70ns. In the meantime, according to ITU-T G.8273.2 amendment 2, Class A and B are extended for Telecom Slave Clock (TSC) as well as T-BC with two more classes, Class C with 30ns requirement and Class D with as low as 5ns.
Further Recommendations
The ITU-T has specified three PTP profile recommendations to be used by telecom operators. ITU-T G.8265.1 “Precision time protocol telecom profile for frequency synchronization” describes telecommunication applications based on IEEE 1588 and defines the PTP profile for unicast mode and frequency distribution.
ITU-T G.8275.1 “Precision time protocol telecom profile for phase/time synchronization with full timing support from the network” specifies the IEEE 1588 functions that are necessary to ensure network element interoperability. This interoperability may be crucial for Mobile Operators for the delivery of accurate phase/time synchronization in a multi-vendor telecom environment. The profile is based on the full-timing support from the network architecture by placing boundary clocks at every node of the network.
ITU-T G.8275.2 “Precision time protocol telecom profile for time/phase synchronization with partial timing support from the network” specifies the ITU-T PTP profile for phase/time distribution with partial timing support from the network. This means usage of both Transparent & Boundary clocks, by placing the boundary clocks at specific network locations to assist in keeping the noise levels as low as possible. This Recommendation defines the PTP profile for unicast mode only, while future editions may contain a separate profile for a mixed unicast/multicast case.
In the meantime, a new version of PTP 1588 is currently under standardization, namely, IEEE 1588-v2.1. This standard is expected to be completed by the end of 2019. It focuses on issues such as new management and performance monitoring, architecture, and protocol security. One of the key focus points is the new High Accuracy profile, based on “White Rabbit,” initially developed by CERN, Geneva, Switzerland.
