Context Augmented Spectrum Sensing in Cognitive Radio Networks
Spectrum management has become a crucial issue in wireless networks. However, perfect utilization of the spectrum among different users is not a trivial task. Over the last two decades, the wireless communication has witnessed significant increase in applications. However, with the persistent evolution of wireless applications, fixed strategies for allocating the spectrum bands cannot handle multiple requirements simultaneously. More importantly, licensed users or primary users in wireless networks are intermittently connected which leading to spectrum underutilization. All of these limitations made it imperative that efficient strategies are developed to manage the spectrum among multiple users or networks. Cognition as a component of intelligence has been employed in communication technologies such as Cognitive Radio Networks for the reasoning and learning goals. From this perspective, CR is an adaptive data network that applies cognition as an optimization tool that aims to optimize the spectrum sharing among multiple secondary users beside the primary users in an autonomous and dynamic way. Spectrum Sensing is one of the most important elements of Cognitive Radio technology since it leads all the subsequent stages of the cognition cycle based on the sensing results. However, with a stand-alone Cognitive Radio devices, local spectrum sensing techniques such as the Energy Detector might draw a false conclusion about the presence of primary transmitter due to several reasons (e.g. fading, shadowing, hidden node problem, noise uncertainty, etc). In cooperative sensing, the uncertainty due to those factors is minimized by exploiting the spatial variation of SUs and concluding one global decision about the PU’s presence/absence. In this research, we propose an intelligent cooperative spectrum sensing where the contextual information of each secondary user is augmented in the fusion process to optimize the final decision of the primary user’s existence. By incorporating the contextual information of the SUs, the spectrum sensing decision could be improved in the sense that the false rejection and false acceptance will be minimized, and therefore interference and utilization will be optimized. Artificial Neural Network as a Machine Learning and Artificial Intelligence tool has been employed as a fusion algorithm utilizing the context of every SU to optimize the final decision. The performance of the proposed technique has surpassed the performance of other fusion algorithms in terms of the sensing reliability and the computational cost.