Abstract
Accurate and precise frequency synchronization is an essential requirement for different areas in telecommunication industry. Emerging practise for frequency synchronization is to utilize packet networks since it is highly cost effective. One method of distributing frequency over a variety of packet networks is based on well known IEEE 1588v2 PTP standard that uses a master-slave architecture. Accuracy and precision of the frequency synchronization over IEEE 1588v2 PTP is mainly deteriorated by the Packet Delay Variations (PDVs) experienced on the transmission path. This problem can be overcome by deploying so-called timing aware routing elements, however slaves' frequency accuracy and precision still heavily relies on the quality of the slave's clock synchronization algorithm. Hence, this work introduces an improved Kalman filter based algorithm for frequency synchronization over IEEE 1588v2 protocol that outperforms other prior art techniques. The algorithm's performance is evaluated using simulation, yet the network impairments (PDVs) are based on experimentally measured data.
Original language | British English |
---|---|
Article number | 7096944 |
Pages (from-to) | 1181-1184 |
Number of pages | 4 |
Journal | IEEE Communications Letters |
Volume | 19 |
Issue number | 7 |
DOIs | |
State | Published - Jul 2015 |
Keywords
- Clock skew
- frequency synchronization
- IEEE 1588
- Kalman filter
- linear programming
- linear regression