Clock synchronization is one of the important issues to be considered in the packet transport network (PTN). Various technologies such as synchronous Ethernet, IEEE 1588v2, and network time protocol (NTP) can be used to achieve clock synchronization. The synchronization status information (SSM) algorithm of the synchronous Ethernet standard has the problems of clock looping and difficulty in tracking and statistics of nodes. ZTE has proposed an extended SSM algorithm to improve clock synchronization. In terms of time synchronization, since the accuracy of NTP cannot meet the requirements of the telecommunications network, using only 1588v2 will bring about problems such as slower convergence time and time delay accuracy when the network load is heavy. ZTE has proposed a 1588v2 time transfer scheme based on synchronous Ethernet, which plays a better role in improving the accuracy of time synchronization in PTN networks.
When operators' demand for packet transmission networks (PTNs) to replace traditional time division multiplexing (TDM) transmission networks is becoming increasingly apparent, how to solve clock synchronization becomes one of the important issues. There are two aspects to the synchronization requirements of the packet transport network: First, it can carry TDM services and provide a mechanism for clock recovery of TDM services, so that TDM services still meet certain performance indicators after passing through the packet network (such as ITU-TG.823 / G .824 specification); Second, the packet network can provide a high-precision network reference clock like a TDM network to meet the synchronization needs of network nodes (such as base stations).
1 Synchronization technology
Clock synchronization includes: frequency synchronization and time synchronization. Frequency synchronization requires the same time interval, and time synchronization requires the same time starting point and the same time interval.
Different wireless technology standards have different requirements for clock bearing. GSM / WCDMA uses asynchronous base station technology, only frequency synchronization is required, and the accuracy requirement is 0.05 ppm, while TD-SCDMA / CDMA2000 requires time synchronization, and TD-SCDMA accuracy requirements It is ± 1.5 μs.
Since 2004, the International Telecommunication Union Telecommunication Standardization Department (ITU-T) Q13 / SG15 has gradually developed a series of recommendations on packet network synchronization technology, mainly including: G.8261 (define the overall demand), G.8262 (define equipment Clock performance), G.8264 (mainly defines the architecture and synchronization function module).
IEEE released the IEEE 1588 standard in 2002, which defines a precise time synchronization protocol (PTP). IEEE1588 is a standard formulated for LAN multicast environments. In the complex environment of telecommunications networks, applications will be limited. Therefore, IEEE1588v2 (hereinafter referred to as 1588v2) was released in 2008, and the technical features adapted to the application of telecommunication networks were added to this version [1-5].
The Internet Engineering Task Force (IETF) Network Time Synchronization Protocol (NTP) implements time synchronization between users and time servers on the Internet.
2 Synchronous Ethernet technology
The physical layer synchronization technology is widely used in the traditional synchronous digital hierarchy (SDH) network. Each node can extract the line clock from the physical link or obtain the clock from the external synchronization interface, select the clock quality from multiple clock sources, lock the local clock to the highest quality clock source, and transmit the locked clock to the downstream device. Through the step-by-step locking, the whole network is synchronized step by step to the main reference clock (PRC). A similar technique can also be adopted for packet networks, the principle of which is shown in Figure 1.
2.1 Principle of Synchronous Ethernet
Synchronous Ethernet technology in packet networks is a technology that uses the Ethernet link code stream to recover the clock. The Ethernet physical layer coding adopts 4B / 5B (FE) and 8B / 10B (GE) technologies. On average, an additional bit is inserted every 4 bits, so that there will not be 4 consecutive 1s in the data stream transmitted by it. Or 4 zeros, which can effectively contain clock information. Use a high-precision clock to send data on the Ethernet source interface, and recover and extract this clock at the receiving end. Clock performance can maintain high precision.
The principle of synchronous Ethernet is shown in Figure 2. In FIG. 2, the sending-side device (node ​​A) injects a high-precision clock into the physical layer chip of Ethernet. The physical-layer chip uses this high-precision clock to send data out. The physical layer chip of the device (node ​​B) on the receiving side can extract this clock from the data stream. In this process, the accuracy of the clock will not be lost, and it can be synchronized with the source to ensure accurate clock. The mechanism of synchronous Ethernet transfer clock is basically similar to that of SDH network. It also recovers the clock from the Ethernet physical link. Therefore, the quality of the recovered clock is not affected by the link traffic. It can provide the same clock tree deployment and SDH / SONET network. The clock quality fully meets the timing interface specifications specified by G.823.
2.2 Synchronous Ethernet SSM algorithm
The Synchronization Status Information (SSM) algorithm is derived from the SDH clock synchronization control, and the usage rules and clock selection algorithm conform to the ITU-T G.781 specification. The SSM control of synchronous Ethernet inherits the SDH network characteristics, and enriches the support of synchronous Ethernet by adding an Ethernet synchronization message channel (ESMC) on the basis of the traditional clock network. It is described in G.8264. The Ethernet synchronization message channel is a unidirectional broadcast protocol channel at the media access control (MAC) layer and is used to transfer synchronization status information SSM between devices. The device selects the optimal clock source based on the SSM information of the ESMC message.
Although the standard SSM algorithm can achieve good synchronization of the network clock, it has two shortcomings: First, it cannot deal with the problem of synchronous clock looping. Special attention needs to be paid in engineering and clock configuration to ensure that clock loops are avoided. The second is the problem of clock signal attenuation. As the number of synchronization links increases, the drift caused by noise and temperature changes in the synchronization allocation process will gradually degrade the quality of the timing reference signal, so the actual number of synchronizable network elements on the same synchronization link is limited Yes, and it is difficult to track and count the nodes through the standard SSM.
ZTE PTN equipment uses an improved extended SSM algorithm, which uses two types-length-value (TLV) to transmit SSM information in ESMC messages. The first TLV transfers the original SSM byte information to the synchronization quality level, following the ITU-T standard; the other TLV is used for path protection. The improved algorithm has the following advantages:
Fundamentally prevents the clock from looping. When there are multiple clock paths, the optimal (shortest) route is automatically selected. As long as there is a route to the main clock, the network element will track the main clock without entering the free oscillation state. The algorithm is low-level distributed processing, so each network element has the same status and simple operation. The standard S1 byte can be used directly without affecting the docking with other manufacturers' equipment.
3 Time synchronization technology
Time synchronization technology is a further development of frequency synchronization. The packet time synchronization technology uses a packet protocol data unit as a carrier of clock or time information, which is a better way to achieve synchronization between the master clock and the slave clock. The basic principle is shown in Figure 3.
3.1 Network Time Protocol
Before the advent of IEEE 1588v2 technology, there were three main protocols used for time synchronization in packet networks: time protocol, time-of-day protocol, and network time protocol (NTP). NTP is implemented by pure software, and the accuracy is relatively low. Currently widely used NTPv3 can achieve synchronization accuracy of about 10 ms. IETF is working on standardization of NTPv4, supports IPv6 and dynamic discovery servers, and is expected to achieve synchronization accuracy of 10 μs. The stability and accuracy of NTP cannot meet the high requirements of telecommunication networks.
3.2 1588v2 protocol
3.2.1 Implementation principle of 1588v2 protocol
1588v2 is a method for providing time synchronization and frequency synchronization in the future. It can be suitable for inter-office time-frequency transmission on different transmission platforms. It can not only transmit frequencies unidirectionally based on the timestamp of 1588v2 in a packet-based time transmission (TOP) method, but also Using the IEEE 1588v2 protocol to achieve time synchronization is widely used in PTN equipment.
The core idea of ​​1588v2 time synchronization is to use the master-slave clock method to encode time information, and use the network symmetry and delay measurement technology to achieve master-slave time synchronization through two-way interaction of message messages.
The principle of the 1588v2 protocol is shown in Figure 4. In the figure, Delay = (T2-T1 + T4-T3) / 2, Offset = (T2-T1-T4 + T3) / 2.
Sync, Follow_Up, Delay_Req, and Delay_Resp messages are sent between the master clock (Master) and the slave clock (Slave). Through the four values ​​T1, T2, T3, and T4, the master-slave time can calculate the delay (Delay) between MaSTer and Slave, and the time difference (Offset) between Master and Slave.
The types of synchronization messages are general messages and event messages. General messages (such as Follow_Up) do not time-stamp themselves, they can carry the accurate sending or receiving time of event messages (such as Sync), and also have the function of completing network configuration, management, or communication between PTP nodes. The event message itself needs to be time-stamped and can carry time stamps or not. The slave clock calculates the time difference between the path delay and the master-slave clock based on the time stamp of the event message or the time stamp carried by the general message.
Wall Mount Digital Signage,Touch Screen Digital Displays,Advertising Led Display Screen,Wall Mount Digital Signage Player
ALLIN , https://www.apiodisplays.com