基于网台分选的目标跳频序列置信区间的预测

Prediction of confidence interval of target frequency hopping sequence based on network station sorting

通信频率的不断变化给非合作通信网络的侦察和预测带来了困难。由于跳频序列通常由具有一定规律性的特定模型生成,因此跳频序列是可预测的。通过提取跳频信号参数,基于时、频二维信息的网台分选算法,研究了基于网台分选的目标跳频序列置信区间预测。利用截获的大量跳频信号数据,构建长短时记忆神经网络模型,用于跳频序列预测。仿真结果表明,即使在一定噪声的影响下,此分选方法依旧可以准确地分选目标跳频信号。此外,基于LSTM的预测方法可以有效地预测跳频序列置信水平和置信区间。

The constant changes in communication frequencies bring difficulties to the reconnaissance and pre-diction of non-cooperative communication networks.Since frequency hopping sequences are usually generated by specific models with certain regularity,frequency hopping sequences are predictable.The target frequency hop-ping sequence confidence interval prediction based on network station selection is studied by extracting frequency hopping signal parameters and a network station selection algorithm based on time and frequency two-dimensional information.Using a large amount of intercepted frequency hopping signal data,a long short-term memory neural network model is constructed for frequency hopping sequence prediction.The simulation results show that even under the influence of certain noise,the sorting method can still accurately sort the target frequency hopping sig-nals.In addition,the LSTM-based prediction method can effectively predict frequency hopping sequence confi-dence levels and confidence intervals.