Electromagnetic Spectrum Occupancy State Volatility Analysis Based on EGARCH Process

In order to describe well the nonlinear time-varying characteristics of spectrum occupancy states which has not been related previously, a novel spectrum occupancy state time series modeling method based on Exponential Generalized Auto Regressive Conditional Heteroskedasticity process (EGARCH) is proposed. Firstly, due to the variance of spectrum occupancy Auto Regressive Moving Average (ARMA) time series model through conditional heteroskedasticity test, it is demonstrated that spectrum occupancy time series has volatility clustering characteristics. Secondly, due to the fitting models analysis results based on EGARCH process and monitoring data, the accuracy of fitting and predicting is better than ARMA model. Thirdly, the leverage coefficients of EGARCH model demonstrate that the influence from spectrum occupancy to electromagnetic environment fluctuation is asymmetric. All above results show that EGARCH model quantifies the complicated nonlinear time varying process of spectrum occupancy.