Statistical service modeling and real-time prediction method for satellite Internet of Things

Abstract

The invention discloses a statistical service modeling and real-time prediction method for the satellite Internet of Things, and the method comprises the following steps: analyzing influence factors, considering the influence of clock drift and terminal electric quantity into service modeling, and obtaining the terminal clock drift amount and the battery electric quantity through a clock drift formula and an ampere-hour integral method; performing statistical service modeling, analyzing service conditions in a typical region, adding influence factors to obtain a regional lumped service volume, and sending the regional lumped service volume to a satellite to obtain an on-satellite service curve; and real-time service prediction is carried out, a parallel CNN-BiLSTM network is constructed, the prediction performance is improved by using an Adaboost algorithm, and real-time prediction of the service volume is completed. According to the method, a traditional 3GPP service model is combined with the Markov Poisson process, actual influence factors are considered, satellite Internet of Things source service modeling is achieved, the working state of each terminal can be reflected, the accuracy of the service model is effectively improved, meanwhile, service prediction is conducted through a CNN-BiLSTM-Adaboost network, and the accuracy of service prediction is effectively improved.

Images