Research On Spectrum Sensing And Power Allocation Technology In Wideband Cognitive Radio Networks

With the rapid development of wireless communication technology, the demand for the wireless spectral resources is larger and larger. Traditional spectral allocation strategy is no longer applicable because of its low utilization, and dynamic spectrum allocation policy can alleviate the shortage problem of spectrum resources to a certain extent, but not fundamentally solve the problem. Accordingly, the shortage of spectrum resources becomes the significant bottleneck of the further development of wireless communication technology in the future. Cognitive radio technology is proposed by scholars in this situation. Spectrum sensing technology, as one of the most important technologies, can dynamically identify the spectrum holes which are not occupied temporarily by primary users(PUs), achieve the spectrum access opportunistically and the reuse of the spectrum resources, thus improving the utilization of spectrum resources. Therefore, there is no doubt that spectrum sensing technology in cognitive radio network becomes the hot topic at this stage.The thesis focuses on investigating the maximization problem of aggregate throughput by jointly optimizing sensing parameters and transmission parameters in wideband cognitive radio network. First, the related theory of spectrum sensing and power allocation in cognitive radio is stated. Then the joint optimization model of spectrum sensing and power allocation in cognitive radio network which is the core research content of the thesis is proposed, that is to say, maximizing the aggregate throughput by jointly optimizing the decision threshold, sensing time and transmission power under the constraints of local interference, false-alarm probability, missed detection probability and maximal total power of each secondary user. It is discovered that the optimization problem can be reformulated into two sub-problems and analyzed respectively: one is the sub-problem of power allocation optimization, an adaptive power allocation based on water-filling is proposed to solve it; the other is the sub-problem of joint optimization of decision threshold and sensing time, theoretical analyses show that the objective optimization function is the unimodal function of sensing time, thus considering using the gold section search method to find the optimal sensing time, moreover, due to the non-convexity of objective function containing the decision threshold, a heuristic Taguchi algorithm is employed to search the optimal decision threshold, therefore, the gold section search method combined with Taguchi algorithm is proposed to solve the optimization problem.The simulation results present the fact that the proposed adaptive power allocation algorithm is more effective than the equal power allocation algorithm, the power allocated by the adaptive power allocation algorithm is corresponding to channel gain-interference ratio, and higher aggregate throughput can be obtained. Under the same condition, the joint optimization algorithm of the gold section search method combined with Taguchi algorithm is more powerful than the Taguchi algorithm, and can get higher aggregate throughput. When the local interference increases, the aggregate throughput is also increasing, but the average optimal sensing time decreases, which really matches with the theoretical analysis; in addition, the values of the aggregate throughput and the average optimal sensing time tend to be stable when the local interference reaches to a certain value.