一种改进的图域频谱感知算法

An Improved Graph-based Spectrum Sensing Algorithm

针对现有基于完全图检测的频谱感知算法存在的计算复杂度高、低信噪比时性能不佳的问题,提出了一种基于分组极差图的改进算法。该算法将观测信号的功率谱以适当的长度进行分组,而后提取其分组极差序列,将其转换成特定无向简单图,并利用图的Gini系数作为完全图特征以实现对信号的检测。相较于现有的图域感知算法,引入分组处理环节能够有效减少图域变换中的处理样本量,且Gini系数计算无需高维矩阵的特征分解,因而计算复杂度降低。仿真结果表明,信噪比大于等于-4 dB时算法检测概率可达0.96以上,平均单次运行时间约为原有图域频谱感知算法的5.6%,有效提高了检测性能并减少了计算复杂度。

Since the existing spectrum sensing algorithms based on complete graph detection have high computational complexity and poor performance at low signal-to-noise ratio(SNR),an improved algorithm based on the graph generated from the range of the block extremes is proposed.The algorithm divides the power spectrum of the observed signal into each block of appropriate length,from which the range of the block extreme sequence is extracted.Accordingly,the sequence is converted into a specific undirected simple graph and its Gini coefficient is used as a complete graph detection feature to detect the signal.Due to introducing the block processing,the sample size of the signal-to-graph transformation is reduced,and the Gini coefficient is calculated without the eigen decomposition of the high-dimensional matrix.When the SNR is greater than or equal to-4 dB,the detection probability of the proposed algorithm can reach more than 0.96 and the average single run time is about 5.6 percent of the existing graph-based spectrum sensing algorithms.The simulation results show that the proposed algorithm effectively improves the detection performance and reduces the computational complexity.