Abstract

Accurate and precise clock synchronization is one of the fundamental requirements for various applications, such as telecommunication systems, measurement and control systems, and smart grid systems. Precision time protocol (PTP) was designed and specified in IEEE 1588 to meet that requirement. PTP provides a mechanism for synchronizing the clocks in a PTP system to a high degree of accuracy and precision based on exchange synchronization messages through a master–slave hierarchy. The best master clock (BMC) algorithm is currently used to establish the master–slave hierarchy for PTP. However, the BMC algorithm does not provide a fast recovery mechanism in case of master failures. The accuracy and precision of the PTP clocks could be deteriorated by the occurrence of failure in the network (link or node failure). These fault occurrences will affect network performance and reliability, and cause clock time drifting of the PTP nodes. In this paper, we present a novel approach, called timing fault recovery (TFR), to significantly reduce clock time drifting in PTP systems. TFR detects the fault occurrence in the network and recovers it by using a handshake mechanism with a short duration. Therefore, the TFR approach provides clock stability and constancy and increases the reliability and the availability of PTP systems. The performance of TFR has been analyzed and compared to that of the standard PTP. Various simulations were conducted to validate the performance analysis. The results show that, for our sample network, the TFR approach reduces clock drifting by 90% in comparison to the standard PTP, thus providing better clock firmness and synchronization accuracy for PTP clocks.

Highlights

  • High-precision clock synchronization has become one of the key requirements for distributed systems in many applications, such as telecommunication systems, measurement and control systems, and smart grid systems

  • The simulation results show that the total delay duration was reduced by about 100% in comparison to the standard precision time protocol (PTP) because the standard PTP does not include a fault detection and recovery mechanism

  • The simulation results show that the timing fault recovery (TFR) approach significantly reduced the local clock drifting of the slave clocks (SCs) in comparison to the standard PTP by about 90% and the standard PTP with rapid spanning tree protocol (RSTP) by about 75%, in the case of fault occurrence between the master clocks (MCs) and the SC

Read more

Summary

Introduction

High-precision clock synchronization has become one of the key requirements for distributed systems in many applications, such as telecommunication systems, measurement and control systems, and smart grid systems. Clocks are synchronized by means of dedicated messages and protocols through communication networks. Two prevalent protocols are used for clock synchronization in these networks: network time protocol (NTP) [1] and precision time protocol (PTP) [2]. NTP is widely used to synchronize computer clocks on the internet. The current version, NTPv4 [3], was formalized in IETF RFCs 5905–5908 in 2010. The primary drawback of NTP is the limitation level of the accuracy of milliseconds, it is cost-effective and it does not require any specific hardware assistance. For distributed computer networks in information technology (IT)

Results
Discussion
Conclusion
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call