A modified ARIMA model for CQI prediction in LTE-based mobile satellite communications

Abstract

LTE-based mobile satellite communication system is a combination of terrestrial LTE system and satellite communication techniques, which aims at improving the current LTE system's technical specifications for better service quality. AMC (Adaptive Modulation and Coding) technique is applied in current LTE system to improve system capacity and data transmission efficiency. CQI (Channel Quality Indicator), which is measured by UE, is an essential indicator for AMC process. However in satellite system, the said data received by eNodeB are all outdated due to the long delay caused by satellite link. So there exists need for predicting CQI data in advance. ARIMA (Autoregressive Integrated Moving Average Model) is a good choice for predicting CQI in terrestrial system, but its prediction ability is limited when prediction length is too long in satellite communication system. To solve this problem, a prediction scheme is proposed in this paper. This method predicts CQI states, instead of predicting the specific CQI values, and provides the reference CQI for AMC. A modified ARIMA model is applied for reducing the prediction complexity and delay. Simulation shows that the proposed method can realize CQI prediction with acceptable RMSE performance.