Abstract:
Cognitive radio essentially depends on optimum spectrum sensing for primary
user detection. Noise uncertainty in spectrum sensing makes the detection process with
fixed threshold unreliable due to thermal noise and interference from other remote communication
systems, which in turn results in variation in the signal to noise ratio (SNR). In
this paper, a dynamic detection threshold under noise uncertainty scheme is proposed for
spectrum sensing to improve the detection performance in an environment characterized
with noise uncertainty and low SNR. Hence, the detection threshold at each secondary user
is dynamically changing according to the predefined detection and false alarm probabilities
together with the received SNR at each node. Furthermore, our proposed integrated
algorithm aims at finding the targeted number of samples, sensing time and user’s
throughput, while maintaining the detection performance metrics within the desired
thresholds. A derived mathematical model and computer simulations are provided to show
the influence of the dynamic threshold on system performance, and proof the robustness of
our proposed scheme under noise uncertainty environment. Our results show a considerable
reduction in number of sensed samples (up to 27%) compared to the approach in
literature under low SNR.
