| 摘要: |
| 针对获取信息空白区域的频谱态势问题，提出了一种无人机（Unmanned Aerial Vehicle,UAV）频谱监测与堆叠生成对抗网络算法结合的频谱地图构建方法。利用无人机的高机动性，快速地对任务区域进行功率谱采样，随后使用得到的采样数据构建完整的频谱地图。该方法能够有效降低获取空域功率谱数据的难度，智能化地对残缺频谱数据进行预测补全，并能根据实际任务需求灵活选择算法堆叠层数。仿真结果表明，30%采样率构建得到的频谱地图与实际频谱地图对比，均方误差低于-20.56 dB。在更低采样率下，堆叠生成对抗网络算法的构建结果相较其他频谱地图算法，误差降低超过-15.66 dB。该算法可用于多种采样方式下的残缺频谱地图预测补全，并构建出与实际态势相符的频谱地图。 |
| 关键词: 频谱态势地图 频谱感知 堆叠生成对抗网络 |
| DOI：10.20079/j.issn.1001-893x.240620007 |
|
| 基金项目:国家自然科学基金资助项目（62071440） |
|
| A Method for Constructing Airspace Spectrum Situation with Multiple UAVs |
| CHEN Yifei,DU Yihang,ZHANG Yu,LONG Weijun |
| (1.School of Electronic and Information Engineering,Nanjing University of Information Science and Technology,Nanjing 210044,China;2.The 63rd Research Institute,National University of Defense Technology,Nanjing 210007,China) |
| Abstract: |
| To address the problem of obtaining the spectrum situation in information gap regions,a spectrum map construction method is proposed.It combines unmanned aerial vehicle(UAV) spectrum monitoring and stacked generative adversarial network algorithm.Leveraging the high mobility of UAVs,the method rapidly samples the power spectrum of the target area,and then uses the obtained sampling data to construct a complete frequency spectrum map.This approach effectively reduces the difficulty of acquiring airspace power spectrum data,intelligently predicts and completes missing spectrum data,and allows flexible selection of the number of algorithm stacking layers according to actual task requirements.Simulation results show that for a 30% sampling rate,the frequency spectrum map constructed has a mean squared error of less than -20.56 dB when compared with the actual spectrum map.At lower sampling rates,the stacked generative adversarial network algorithm results in an error reduction of over -15.66 dB compared with other spectrum mapping algorithms.The algorithm can be used for predicting and completing incomplete spectrum maps under various sampling methods,and constructing frequency spectrum maps that align with the actual situation. |
| Key words: spectrum situation map:spectrum sensing:stacked generative adversarial network |
《电讯技术》编辑部 