Spectrum Sensing in Full-Duplex Cognitive Radio Networks under Hardware Imperfections



Direct-conversion radio transceivers can offer re-programmable and low-cost hardware solutions for full-duplex (FD) cognitive radio networks (CRNs). However, they are susceptible to radio frequency (RF) impairments, such as in-phase(I) and quadrature (Q) imbalance (IQI), which can significantly limit the spectrum sensing capabilities. This paper is devoted to quantify and evaluate the effects of IQI in single- and multi-channel energy detectors operating in FD mode, under both cooperative and non-cooperative spectrum sensing scenarios. In this context, closed-form expressions are derived for the false alarm and detection probabilities in the general case, where partial self-interference suppression (SIS) and joint transmitter(TX) and receiver (RX) IQI, are considered. Furthermore,simplified closed-form expressions for the special cases, where either the RF front-end is ideal or the SIS technique is perfect,are also presented. The presented analytical results have been verified through extensive simulations and indicate that the IQI and partial SIS can significantly affect spectrum sensing accuracy in FD-based CRNs. Specifically, if ideal RF front-end is assumed,spectrum sensing error can significantly increase, leading to a reduction in the CRN performance and negatively affect the performance of primary (PR) networks. Hence, when designing spectrum sharing algorithms for FD-based CRNs, the hardware impairments should be considered in order to improve the CRN performance, while minimizing the negative effects on PR users.