抗干扰性能评估是构建短波跳频通信系统的关键环节。给出了一种基于神经网络的评估方法。该方法有如下优点 :为了得到可供参照的数据 ,利用 Delphi法和 AHP法相结合确定一组学习数据 ;提出了一种新的带“奖优罚劣”性质的原始数据处理方法 ;利用神经网络的自学习功能 ,适应复杂多变的战争环境 ;神经网络可用简单硬件实现 ,简单可靠 Anti-jamming performance evaluation is the key to build shortwave frequency hopping communication system. An evaluation method based on neural network is given. The method has the following advantages: In order to obtain the data for reference, a set of learning data is determined by combining the Delphi method and the AHP method; a new original data processing method with “good and bad” is proposed; Neural network self-learning function, to adapt to the complex and ever-changing war environment; neural network can be simple hardware to achieve, simple and reliable