What is Time & Frequency Network Intelligent Networking Technology?

What is Time & Frequency Network Intelligent Networking Technology?

Time synchronization has been the first need for human information interaction since ancient times. With the rapid development of information technology and communication technology today, various systems have increasingly stringent requirements for time and frequency synchronization. For example, 5G communication networks now require clock sources, clock transmission, and end-level clock synchronization performance to reach the sub-nanosecond level. High-reliability and high-precision time synchronization directly relates to the quality and transmission rate of 5G communication networks. According to clause 9.6.3.3 of 3GPP TS 38.104 V18.8.0 (2024-12), in 5G wireless communication technology, “in-band non-contiguous carrier aggregation” requires a synchronization accuracy of 260ns, “in-band contiguous carrier aggregation” requires a synchronization accuracy of 130ns, and for 480kHz and 960kHz high-frequency base stations, “in-band contiguous carrier aggregation” requires a synchronization accuracy of 32.5ns. Once the synchronization system fails, it will directly lead to the unavailability of related 5G services.

Since high-precision time synchronization networks are crucial for 5G services, major operators currently generally adopt a multi-source time scheme. Each device switches between various time sources based on manually preset priorities to maximize the availability of time sources. This approach does not take into account dynamic factors such as the average failure rate of each time source, which may lead to overly frequent switching of the time synchronization link, thereby affecting the stability of the time synchronization network. For example, a time-synchronized device has three time sources: S1, S2, and S3, with preset priorities of 1, 2, and 3 (1 being the highest priority, 2 next, and 3 the lowest). However, during a certain period, if S1 has an exceptionally high failure rate, the time-synchronized device will switch to S2 when S1 fails, and then switch back to S1 once S1 is restored, resulting in frequent switching phenomena that affect the stability of the time synchronization network.

Our company has designed a brand-new patented technology that adopts intelligent learning techniques. It real-time statistics the annual average downtime of each time source, and combines the time source’s accuracy and stability indicators. Through internal algorithms, it calculates the comprehensive priority of the time source, using this comprehensive priority as the selection standard. This ensures that the receiving timing devices always track the time source with the optimal comprehensive priority such as failure rate, accuracy, and stability, thereby improving the stability of the timing system. This solution can address the phenomenon where, when preset priority methods are used to switch time sources, high-priority time sources frequently experience failures and recover, leading to frequent switching between high and low priority time sources on the time server, thus enhancing the stability of the time server.

The principle block diagram of the time server with intelligent networking is as follows:

After S1, S2, and S3 time sources connected, they is an intelligent source selection module calculate the comprehensive priority, the time source with low failure rate, high accuracy, and high stability is finally selected for use by the synthesis module. This patent technology statistically utilizes the annual average available time, avoiding the frequent switching problems that may arise from the current preset priority-based source selection method, thereby improving system availability and stability. For fiber optic time synchronization systems requiring large-scale networking across the country, this technology can serve as a basic algorithm to form an nationwide intelligent timing network.

In addition, the time server has flexible networking capabilities. The aforementioned S1, S2, and S3 three (or more) time sources can be any type of input time source supported by the time server, such as ground PTP input time sources, 1PPS+TOD time sources, ground Beidou fiber time sources, GNSS satellite time sources, etc.

If you have any questions, please feel free to consult via email or phone.

Email: liyan@groundbeidou.com

Phone: +86 15928828658

Leave a Reply

Your email address will not be published. Required fields are marked *