Kalman Filter Clock Offset and Skew Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing clock synchronization methods over packet networks, such as those using IEEE 1588 Precision Time Protocol (PTP), often assume a clock difference only by offset, neglecting the inherent skew, leading to inaccurate synchronization, especially in critical applications requiring stringent accuracy.
Innovation Solution
A method employing a Kalman filter for recursive estimation of clock offset and skew using IEEE 1588 PTP messages, where timestamps from Sync and Delay_Req messages are processed to accurately estimate and adjust the local clock's offset and skew relative to a remote master clock, leveraging zero-mean Gaussian white noise and process noise vectors within a two-state dynamic model.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If clock synchronization methods assume only offset difference between clocks, then the algorithm complexity is reduced, but the synchronization accuracy deteriorates due to neglecting inherent skew
Solution Approach 1:
The patent changes the model parameters from offset-only to offset-plus-skew, accurately representing the physical reality of clock behavior. By incorporating the skew parameter α into the clock model θ(t) = θ(0) + αt, the system achieves higher synchronization accuracy while managing complexity through efficient estimation algorithms.
Solution Approach 2:
The patent replaces traditional batch processing methods with a recursive Kalman filter approach. This substitution enables continuous real-time estimation of offset and skew parameters, improving synchronization accuracy without proportionally increasing system complexity through incremental updates rather than complete recalculations.
2Measurement precision
If traditional batch processing methods are used for offset and skew estimation, then the estimation accuracy can be maintained, but the computational complexity increases significantly in dynamic network conditions
Solution Approach 1:
The patent implements a dynamic recursive estimation approach where the Kalman filter continuously adapts to changing network conditions. The filter updates offset and skew estimates in real-time as new timestamp measurements arrive, allowing the system to track time-varying clock behavior without requiring complete batch reprocessing, thus reducing computational complexity in dynamic environments.
Solution Approach 2:
The patent establishes continuous estimation through the Kalman filter that processes timestamp measurements as they arrive from periodic PTP messages. This continuous action maintains up-to-date offset and skew estimates without interruption, ensuring accurate synchronization while avoiding the computational burden of periodic batch processing of all historical data.
Data Source
Figure 1
Figure 2a~2b
Figure 3
AI summary
This invention relates to methods and devices for clock offset and skew estimation. The invention has particular application in the alignment of slave clocks to a master clock. In embodiments of the invention, the slave clock employs an independent free running clock and a recursive estimation technique to estimate the clock offset and clock skew between the slave and master clocks. The slave can then use the offset and skew to correct the free running clock to reflect an accurate image of the master clock.