Wireless spectrum prediction model based on time series analysis method

Cognitive Radio (CR) is considered to be a solution for mitigating the conflict between limited spectrum resources and increasing demand of various applications. For the purpose of guaranteeing primary users, we propose a spectrum prediction model using the spectrum historic utilization information with time series analysis. We conduct experiment on wide TV bands of China from 603.25 MHz to 843.25 MHz, the results show that this model can accurately predict status of channels and find vacant spectrum for cognitive users. We take channel 2 whose center frequency locates in 615.25 MHz for example, average forecasting error ratio is 1.01% with a minimum of 0.79%.