基于特征值的频谱感知算法仿真研究

频谱感知是认知无线网络中的一个重要功能,是实现其他功能的基础。研究了新型的基于特征值的感知算法,充分利用随机矩阵的渐近谱分布特性和特征值的收敛特性来设置判决门限,提高感知性能。理论分析和仿真结果均表明,新型算法性能明显优于典型的能量检测算法,克服了能量检测算法的噪声不确定性问题。

Abstract：

Spectrum Sensing is an important function in the cognitive radio networks.A new eigenvalue-based detection scheme for spectrum sensing was studied,and decision thresholds were set to have good performance by using the property of asymptotic spectrum distribution and convergence of maximum eigenvalue of random matrices.Theoretical analysis and simulations results show that the new detection scheme can obviously outperform the classical energy detection,and overcome the noise uncertainty problem.

Simulation and Analysis of Eigenvalue-based Spectrum Sensing Algorithms for Cognitive Radio Networks

Spectrum Sensing is an important function in the cognitive radio networks.A new eigenvalue-based detection scheme for spectrum sensing was studied,and decision thresholds were set to have good performance by using the property of asymptotic spectrum distribution and convergence of maximum eigenvalue of random matrices.Theoretical analysis and simulations results show that the new detection scheme can obviously outperform the classical energy detection,and overcome the noise uncertainty problem.