Forecasting models for road-vehicle communication traffic in ITS

Since the telecommunication system will become one of important keys in an intelligent transport system (ITS), it is indispensable to consider and understand characteristics of the communication traffic to realize an efficient ITS communication network. In addition, forecasts of the communication traffic volume lead to efficient utilization of network equipment and frequency resources. Communication traffic characteristics in ITS networks are different from existing networks in that communication traffic is generated in vehicles. Therefore, we need to create suitable models for this purpose. This paper proposes effective and suitable forecasting models for ITS communication network planning. Since there is no available communication traffic data in ITS, models for generation of communication traffic are also described as well as the vehicle movement models. This paper also introduces a simulator developed based on these models. The proposed forcasting models are evaluated in terms of forecast error by the simulator. It is found from the simulation results that multiple linear regression models using traffic density data give more accurate forecast values than autoregressive models.