Time-triggered Ethernet (TTE) guarantees real-time transmission of communication data through precise time scheduling
and stands out in many fields of data transmission with high real-time requirements. Analyze the TTE synchronous characteristic we can compensate the local clock of synchronous main node through the least square clock compensation algorithm in the project and improve determinacy and real-time nature. However
the local clock of the synchronous master node is vulnerable to the influence of the external environment to generate outliers
which affects the applicability of the least squares clock compensation algorithm. Therefore
this paper proposes a clock compensation algorithm with outlier pre-elimination to improve the real-time performance of the system. That is
firstly
the Least Square criterion (Pauta criterion) is used to pre-eliminate the least-squares-sensitivity local clock outliers
then the local clock compensation value is estimated by least-squares method
and finally the local clock is processed according to the estimated clock compensation value. Simulation results show that the proposed method solves the problem of local outliers occurring at the synchronous master node in the TTE clock compensation. The results show that the clock synchronization accuracy of the proposed algorithm can reach 10 seconds when the outlier ratio is less than 10%. It is much higher than the 10 second-level accuracy of the least squares clock compensation algorithm with no outlier pre-elimination.
Lemmon M D, Ganguly J, Xia L. Model-based clock synchronization in networks with drifting clocks[C]// Pacific Rim International Symposium on Dependable Computing. Washington, DC, USA, IEEE Computer Society, 2000: 177.
Steiner W, Dutertre B. Layered diagnosis and clock-rate correction for the TTE thernet clock synchronization protocol[C]//IEEE, Pacific Rim International Symposium on Dependable Computing. IEEE Computer Society. Washington, DC, USA, 2011: 244-253.
Rousseeuw P. Least median of squares regression[J]. Publications of the American Statistical Association, 1984, 79(388):871-880. DOI: 871
De l T F. A least-squares framework for component analysis[J]. IEEE Transactions on Pattern Analysis % Machine Intelligence, 2012, 34(6): 1041-1055. DOI: 1041
Hong Li H E, Ren P Z, Tian W F, et al. Studty on revised MT filter-cleaner in flight test data processing[J]. Measurement % Control Technology, 2006, 25(4):11-13. DOI: 11