Study On Spectrum Sensing Techniques In Cognitive Radio

With the rapid development of wireless communications technologies and applications, the demand for wireless spectrum resources increases. In the framework of the existing spectrum management, the available spectrum resources that can be allocated become more and more scarce. According to Federal Communications Commission's study report, the assigned spectrum is not being fully used at a specific time and geographic lacation, that is, spectrum resources are underutilized. With the help of cognitive radio technology, cognitive users can dynamically take advantage of these idle spectrum resources for new wireless applications. As a result, the spectrum efficiency can be improved significantly. Moreover, it can effectively alleviate the conflict between the shortage of spectrum resources and the growing demand for wireless access. Therefore, cognitive radio technology has become a hotspot in the research of wireless communications technologies.Cognitive radio system is defined as an intelligent wireless coommunication system that is aware of its environment. Through continuous sensing the changes of the outside world, it can adaptively adjust its internal communication mechanisms to achieve adaptation to the environmental changes. In cognitive radio, the key technologies include spectrum sensing, spectrum management, spectrum sharing, spectrum mobility, and so on. In this dissertation, we focus on the key technology of spectrum sensing, and some problems about energy detection scheme are studied, and the main contributions are as follows:1) In cooperative spectrum sensing, the effects of decision thereshold on the detection performance of local energy detection and cooperative spectrum senseing are studied. A method for setting the optimal decision thereshold for energy detection is proposed and the search region is also given, within which the optimal decision threshold can be quickly obtained. Due to the noise, fading and other effects, the detection performance of a single cognitive user can not meet the performance requirements. Thus, spectrum sensing is usually executed by multiple cognitive users in a cooperative manner. The decision threshold plays a fundmantal role in the performance of local energy detection, and it also affects the performance of cooperative spectrum sensing. In order to minimize the total error probability which includes the false alarm probability and the miss detection probability, the optimal decision threshold for a single cognitive user must be set. The dissertation gives the search region within which the optimal decision threshold can be quickly searched under the given conditions. Simulation results are presented to verify the proposed method. 2) The impacts of noise uncertainty on the performance of spectrum sensing are studied. Simulation results show that the noise uncertainty has seriously reduced the detection performance of energy detection both for a single user or multiuser cooperative spectrum sensing. On this basis, using an estimated noise power to set the decision thresholds for energy detection is proposed. It assumes that the noise power is unknown and needs to be estimated from the reference bands. Then, the problem of calculating the decision thresholds for energy detection when using the estimated noise power is analyzed. The corresponding closed-form detection performances are derived. To meet the expected detection performance, the decision thresholds can't be derived just by replacing the exact noise power with the estimated one, and they must be modified. The closed-form expressions of the modified decision thresholds are given, which greatly simplify the analysis of the detection performances and the calculations of the decision thresholds. Moreover, the closed-form expressions of the detection performance based on the modified decision thresholds are given. Simulation results show that, under the constraint of the expected detection probability, the false alarm probability based on the modified decision thresold is much smaller than that based on the oriented decision threhold. It means that based on the modified decision threshold, the throughput of cognitive radio system can be effectively increased.3) In cognitive radio, a cognitive user can achieve spatial diversity by deploying with multiple receiver antennas, and the reliability of spectrum sensing can be improved. When the multiple antennas are independent each other, the maximum diversity gain can be achieved. In this dissertation, the scenario that the multiple antennas are correlated is considered, and the effects of the antenna correlation on the actual performance (including the sensing efficiency and the sensing accuracy) improvement when the reciever deployed with multiple antennas are investigated. The upper and lower bounds of the false alarm probability with considering the antenna correlation are derived. It has been mathematically proven that, when the signal to noise ratio is low, the cognitive user with correlated antennas can achieve approximately the same actual performance improvement as that with independent antennas. Simulation results verify the correctness of our analysis. Moreover, the results also show that in the regime of low SNR, cognitive user needs more detection time to achieve the maximum actual performance improvement when deployed with more receiver antennas. It is due to the inevitable tradeoff between the sensing efficiency and the sensing accuracy.