Huigang Li / Guilin University of Electronic Technology,
Network delay jitter variation in Cyber-Physical Systems (CPSs) are one of the main factors which cause performance degradation or even system instability, and how to smooth the network delay jitter or reduce the uncertainty is one of several major problems that CPSs face. Based on the analysis of the research status of network delay jitter, a network jitter smoothing algorithm- Adaptive Exponential Smoothing Prediction Algorithm (AESP algorithm) was proposed. This innovation lies in introducing innovative "tracking signal" to track delay variation characteristics. By tracking signal to adjust the error of forecasts and further optimizing the predictive model, the prediction model is made to have the automatic response to this change and re-forecast predicted values. Experiments show that the algorithm has some advantages. The improved algorithm can better predict changes in the size of network delay jitter, which will not only effectively reduce the exceptional value of changing network jitter, reduce the dropout rate, but also has no need for the synchronization of each system module clock .