Improved Power Allocation in Cognitive Radio Networks Under Spectrum Sensing Uncertainties

We propose a power allocation scheme for OFDM-based cognitive radio systems by taking into account the realistic scenario of imperfect spectrum sensing. To this end, we first derive the interference level imposed from the primary network on the cognitive network which has an imperfect spectrum sensing. Then, by using these interference parameters, we maximize the downlink transmission capacity achieved by the cognitive user while keeping the interference level at the primary user (PU) below a predefined threshold. Our numerical results indicate that for a given interference threshold tolerated at the PU, the proposed power loading scheme increases the cognitive achievable data rates compared to classical loading algorithms that consider a perfect spectrum sensing.