Secondary Clock Synchronization Using Adaptive Kalman Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing clock synchronization methods, such as linear and extended Kalman filters, struggle to respond quickly and robustly to sudden frequency changes in local oscillators, leading to instability or the need for complex state machines.
Innovation Solution
A synchronization technique using a Kalman filter with adjustable offset and skew variances to account for internal noise and external influences, enabling fast and robust frequency stabilization of secondary clocks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If extended Kalman filters are used to handle rate ramp, then the ability to follow sudden frequency changes is improved, but instability problems occur when the filter is used outside the vicinity of the operating point
Solution Approach 1:
The patent changes the parameters of the Kalman filter by introducing time-varying process noise covariance matrices that adapt to different operating conditions. This allows the filter to maintain stability while responding to frequency changes by adjusting its parameters rather than using a complex extended Kalman filter structure.
Solution Approach 2:
The patent makes the Kalman filter dynamic by allowing the process noise covariance to vary over time based on the detected operating condition. This dynamic adaptation enables the filter to handle both steady-state and transient conditions effectively without sacrificing stability.
2Adaptability or versatility
If Sage-Husa Kalman filters are used to determine rate ramp adaptively, then the flexibility of the model is improved, but the response time to sudden frequency changes becomes too long
Solution Approach 1:
The patent performs preliminary classification of the operating condition (steady-state or transient) before applying the appropriate filtering strategy. This preliminary action allows the system to quickly respond to frequency changes by immediately switching to the appropriate mode without the delays associated with adaptive parameter tuning.
Solution Approach 2:
The patent segments the operation into distinct modes (steady-state and transient) and applies different filtering strategies to each segment. This segmentation allows for optimized performance in each mode while maintaining overall system responsiveness.
3Adaptability or versatility
If linear Kalman filters are extended with state machines to detect changing circumstances, then the ability to handle different situations is improved, but the implementation complexity increases and leads to sub-optimal solutions
Solution Approach 1:
The patent enables the Kalman filter to self-adapt by using the residual error from the filtering process to detect transient conditions and automatically adjust its behavior. This self-service mechanism eliminates the need for external state machines while maintaining the ability to handle different operating situations optimally.
Solution Approach 2:
The patent implements feedback by using the filter residual to detect operating conditions and adjust the process noise covariance accordingly. This feedback loop provides adaptive behavior without the complexity of state machines, as the system automatically responds to changing circumstances based on its own performance metrics.
4Device complexity
If rate ramp is modeled as a deterministic process in linear Kalman filter, then the simplicity of the linear filter is maintained, but the rate ramp must be present in the system all the time reducing flexibility
Solution Approach 1:
The patent makes the linear Kalman filter dynamic by allowing the process noise covariance to vary over time based on detected operating conditions. This dynamic approach maintains the simplicity of the linear filter structure while adding the flexibility to adapt to different situations, including the presence or absence of rate ramp.
Solution Approach 2:
The patent changes the parameters of the linear Kalman filter (specifically the process noise covariance) to adapt to different operating conditions. This allows the simple linear filter to maintain flexibility by adjusting its parameters rather than requiring complex model structures or continuous rate ramp presence.
Data Source
AI summary
A technique for synchronizing a secondary clock (100; 800; 900; 1091; 1092; 1130) with a primary clock (300) is provided. As to a first method aspect, a method comprises a step of receiving (202), from the primary clock (300), a synchronization signal (302) indicative of a reference time at each of a plurality of synchronization events. The method further comprises a step of estimating (204) an offset and a skew of the secondary clock (100; 800; 900; 1091; 1092; 1130) based on the reference time and a local time retrieved from the secondary clock (100; 800; 900; 1091; 1092; 1130) for each of the synchronization events. The offset and the skew of the secondary clock (100; 800; 900; 1091; 1092; 1130) are estimated (204) using a Kalman filter, KF, comprising an offset variance and a skew variance. The offset variance being indicative of a power of noise of the offset and the skew variance being indicative of a power of noise of the skew. The offset variance is set according to an internal noise power measured for the secondary clock (100; 800; 900; 1091; 1092; 1130). The skew variance is set according to an external influence on the secondary clock (100; 800; 900; 1091; 1092; 1130) measured for an environment of the secondary clock (100; 800; 900; 1091; 1092; 1130). The method further comprises a step of updating (206) the secondary clock (100; 800; 900; 1091; 1092; 1130) based on the estimated (204) offset and skew.