软件通信适配器的调制模式识别算法

发布时间：2018-06-03 08:35

本文选题：异构卫星网络 + 软件通信适配器　；参考：《东南大学学报(自然科学版)》2017年03期

【摘要】：在异构卫星网络动态组网时,为了解决星上软件通信适配器对物理层调制模式识别率低的问题,提出了一种适合低信噪比和贫先验知识的自动调制模式识别算法.该算法以高斯白噪声信道作为信道模型,选取信号高阶累积量和经典统计量作为特征参数,采用引力搜索算法对径向基神经网络基函数中心进行优化,并在引力搜索算法中引入粒子群的信息熵来调节算法执行过程中探索与开采的关系,进一步提高了算法的分类和泛化能力.然后,利用仿真试验测评了该算法对6种卫星常用调相调制信号的识别效果.仿真试验结果表明,没有先验知识的情况下,该算法在调制信号信噪比大于4 d B时就可以达到100%的识别率,从而证明了该算法在低信噪比和贫先验知识条件下的有效性,说明算法满足星上软件通信适配器对物理层调制模式的识别要求.
[Abstract]:In order to solve the problem of low recognition rate of modulation pattern in physical layer by on-board software communication adapter, an automatic modulation pattern recognition algorithm suitable for low SNR and prior knowledge is proposed. In this algorithm, the Gao Si white noise channel is used as the channel model, the high order cumulant and the classical statistics are selected as the characteristic parameters, and the radial basis function center of the radial basis function neural network is optimized by gravity search algorithm. The information entropy of particle swarm optimization is introduced into the gravitational search algorithm to adjust the relationship between exploration and mining during the execution of the algorithm, which further improves the classification and generalization ability of the algorithm. Then, the recognition effect of this algorithm on six kinds of satellite phase modulation signals is evaluated by simulation test. The simulation results show that without prior knowledge, the recognition rate of the algorithm can reach 100% when the SNR of modulated signal is greater than 4dB, which proves the validity of the algorithm under the condition of low SNR and poor prior knowledge. The algorithm meets the requirements of on-board software communication adapters for physical layer modulation pattern recognition.
【作者单位】：解放军理工大学气象海洋学院;解放军理工大学指挥信息系统学院;
【基金】：国家自然科学基金资助项目(61371119)
【分类号】：TN927.2

【相似文献】