基于改进离散傅里叶变换滤波器组的两步检测算法

Two-step Detection Algorithm Based on MDFT Modulated Filter Bank

认知无线电(cognitive radio,CR)在解决频谱资源匮乏问题和提高频谱资源利用率方面被寄予厚望。频谱感知的循环特征检测算法复杂度较高,而能量检测算法在信噪比比较低的情况下检测性能不理想;针对这种情况,在改进离散傅里叶变换(MDFT)滤波器组结构的多载波调制系统中,提出一种基于MDFT滤波器组调制的两步检测算法。该算法结合了循环特征检测较强的抗噪声干扰特性,以及能量检测算法复杂度低的优势。仿真结果表明:所给出的两步检测算法在较低信噪比时相对于MDFT能量检测、OFDM循环特征检测等其他算法,其检测性能有明显改善;同时MDFT循环特征检测算法可有效降低算法复杂度。

Cognitive radio（ CR） is considered as an important technology which can resolve the problem of spectrum scarcity and improve resource utilization. Cyclostationary detection algorithm of the spectrum sensing had the high complexity,and the performance of the energy detection algorithm in low signal to noise ratio was not ideal. In view of this case,in the Modified Discrete Fourier Transform filter banks-based multicarrier modulation system,a two-step detection algorithm had been presented. This algorithm combined the strong anti-noise characteristics of cyclostationary detection and the low complexity of energy detection. Simulation results show that comparing with MDFT Energy Detection,OFDM Cyclostationary detection and other algorithms,the proposed two-step detection algorithm improve the detecting performance in the low SNR situation. Meanwhile,it can effectively reduce the complexity of the algorithm.