Adaptive Time-of-Day Synchronization in Packet Networks

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing clock synchronization methods in packet processing networks face challenges in accurately estimating frequency skew between master and slave clocks, especially in non-stationary network environments with varying packet delay variations, which can lead to synchronization errors and require frequent recalibration.

Innovation Solution

The method involves accumulating timestamps to determine packet delay variation sequences, adjusting phase offsets using a time-of-day estimation algorithm, and selecting appropriate statistical metrics such as mean, mode, or standard deviation to synchronize clocks, while adapting to changing network conditions by periodically checking for stationarity and updating algorithms accordingly.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If packet-based synchronization methods are used in non-stationary network environments, then clock synchronization can be achieved, but synchronization accuracy deteriorates due to varying packet delay variations

Engineering Contradiction:
Improveclock synchronizationVSAvoidsynchronization accuracy
Core Design Contradiction:
ReliabilityVSMeasurement precision

Solution Approach 1:

The patent applies dynamics by making the synchronization system adaptive to changing network conditions. The slave device periodically checks for stationarity of packet delay variation sequences and switches between different estimation algorithms (stationary vs. non-stationary) based on current network conditions. This dynamic adaptation allows the system to maintain synchronization accuracy despite varying PDV characteristics in non-stationary environments.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The patent changes the parameter of algorithm selection based on network conditions. When the PDV sequence is determined to be stationary, a stationary estimation algorithm is used; when non-stationary, a non-stationary algorithm is employed. This parameter change (algorithm selection) directly addresses the contradiction by matching the estimation method to the current network state, thereby maintaining synchronization accuracy across different conditions.

Inventive Principle:
Principle #35Parameter changes

2Measurement precision

If frequent recalibration is performed to maintain synchronization accuracy, then timing errors are reduced, but network traffic and processing overhead increase

Engineering Contradiction:
Improvetiming errorVSAvoidnetwork traffic
Core Design Contradiction:
Measurement precisionVSQuantity of substance

Solution Approach 1:

The patent implements periodic action through the slave device periodically checking for stationarity of the PDV sequence at predetermined intervals. This periodic monitoring allows the system to adapt to changing network conditions only when necessary, rather than continuously recalibrating. The periodic stationarity checks trigger algorithm switches only when actual changes in network behavior are detected, reducing unnecessary recalibration traffic while maintaining synchronization accuracy.

Inventive Principle:
Principle #19Periodic action

3Measurement precision

If multiple exchange rounds are used for accurate frequency and phase estimation, then synchronization precision is improved, but time required for synchronization increases

Engineering Contradiction:
Improvefrequency and phase estimation accuracyVSAvoidsynchronization time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent uses feedback by having the slave device monitor the stationarity of the PDV sequence and use this information to select the appropriate estimation algorithm. This feedback mechanism allows the system to adapt its estimation approach based on observed network conditions, achieving accurate frequency and phase estimation without requiring excessive exchange rounds. The feedback-driven algorithm selection optimizes the trade-off between estimation accuracy and synchronization time.

Inventive Principle:
Principle #23Feedback

Data Source

PatentUS9444566B1Methods of performing time-of-day synchronization in packet processing networks
Publication Date: 2016.09.13 SITIME CORP
  • US9444566B1 patent drawing
  • US9444566B1 patent drawing
  • US9444566B1 patent drawing

AI summary

Methods of performing time-of-day synchronization in a packet processing network include accumulating timestamps transmitted in packets between master and slave devices, which are synchronized to respective master and slave clocks and separated from each other by the packet processing network. Operations are also performed to determine whether first timestamps accumulated in a first direction across the packet network demonstrate that a first packet delay variation (PDV) sequence observed from the first timestamps is stationary. A phase offset between the master and slave clocks is then adjusted using a time-of-day (ToD) estimation algorithm. This adjusting can include evaluating a location-dependent statistic of the first PDV sequence.