Digital Twin Based Network Latency Prediction in Vehicular Networks

Network latency is a crucial factor affecting the quality of communications networks due to irregularity vehicular traffic. To address problem performance degradation or instability caused by in networks, this paper proposes time delay prediction algorithm, which digital twin technology employed obtain large quantity actual data for and verify autocorrelation. Subsequently, meet conditions ARMA series model, two neural i.e., Radial basis function (RBF) Elman were construct model. The experimental results show that average relative error RBF 7.6%, whereas Elman-NN 14.2%. This indicates has better performance, real-time than Elman-NN.