Limited Data-driven Router Bandwidth Configuration for Cyber Physical Internet

In the Cyber Physical Internet (CPI), router bandwidth reflects the router's ability to process and transmit data. Configuring router bandwidth is crucial to meeting throughput demands, making the evaluation of router throughput a critical factor in bandwidth configuration. This paper investigates the configuration of CPI router bandwidth when data accessibility is limited. The paper models the variation of router throughput using the Autoregressive Integrated Moving Average (ARIMA)-Kalman hybrid evaluation method for bandwidth configuration. The freight data from January 2020 to December 2021 at the rail-water intermodal station were analyzed and processed using the ARIMA-Kalman model for bandwidth configuration. The results indicate that the proposed model can more accurately represent the fluctuation of demand for empty rail-cars and achieve short-term bandwidth configuration within error tolerance and the limited observation dataset.