Practical Energy Detection for Internet of Things Devices

摘要

This work investigates the impacts of unknown parameters of noise and signal powers on the popular spectrum sensing scheme, i.e., energy detection, for cognitive radios (CRs) over fading channels, which is a promising communication technology for the Internet of Things (IoT) devices. To study the effects of unknown parameters for the energy detector, a new maximum-likelihood (ML) estimation of noise and signal powers employing the cyclic prefix (CP) of orthogonal frequency-division multiplexing (OFDM) is presented. The Cramer-Rao lower bounds (CRLBs) of the estimation are obtained. Furthermore, the performances under both hypotheses, i.e., false-alarm rate (FAR) and detection probability (DP), of the impaired energy detector are validated by both simulation and analytical results.