基于时间序列分析的频谱异常自主检测和稳健估计方法

复杂电磁环境和无用频先验知识条件下有效检测电磁频谱异常使用信息，是无线电监测和电磁环境评估等领域的重要难题.本文基于时间序列分析理论，通过构建反映有限频谱占用度序列动态依存关系且包含频谱异常值的时序模型，实现对无线电频谱异常的自主检测和稳健估计.研究结果表明，该方法无需用频数据库或无线电监测历史数据支持，能够有效识别典型频谱异常类型、发生时间以及异常影响强度等信息；同时通过对频谱占用度时序模型的稳健估计，能够显著降低模型拟合误差，提高模型对外部干扰环境的适应性和鲁棒性.

An Autonomous Detection and Robust Estimation Method of Spectrum Anomaly Based on Time Series Analysis

How to precisely detect electromagnetic spectrum anomaly is a major challenge for radio monitoring and electromagnetic environment evaluation,especially in the condition of complex electromagnetic environment and lack of pre-knowledge information about frequency use.Based on time series analysis theory,a timing model which presents the correlation between the previous and the following sequence of spectrum occupancy,is built to help us realize the autonomous identification and robust estimation of typical spectrum anomaly.The analysis results indicate that,without actually requiring pre-knowledge of frequency database and radio monitoring historical data support,this method can effectively identify the types of spectrum anomaly,occurrence time,anomaly effect intension and other relative information.Furthermore,through the robust estimation of spectrum occupancy model,we can significantly improve the model's fitting performance and raise the adaptability and robustness of the model to external interferences.