Queue-Aware Spectrum Sensing for Interference-Constrained Transmissions in Cognitive Radio Networks

We propose the queue-aware spectrum sensing schemes for interference-constrained opportunistic transmissions of secondary users (SUs) in cognitive radio networks. Specifically, we employ the energy detection for SUs to sense the spectrum-usage status of the primary users (PUs). Unlike the traditional energy detector using a fixed energy threshold to decide the presence of the PUs' signals, we dynamically change the energy threshold based on the queue length at the sender of SUs. Our proposed dynamic-threshold policies aim at effectively satisfying the statistical quality-of-service (QoS) requirements such as the queue-length-bound violation probability and buffer-overflow probability, while upper-bounding the probability of causing interferences to the PUs. We develop the dynamic-threshold policies for the scenarios that the sender of SUs has infinite and finite queue buffer sizes, respectively. Simulations evaluations show that under the specified statistical QoS requirements and interference constraints, our proposed schemes can support higher data traffic loads for SUs than the traditional energy-detection based scheme.