基于深度神经网络的移动通信网络优化研究

Mobile communication network optimization based on deep neural network

针对移动通信网络优化过程中异常信号的识别控制较为困难,而传统的异常信号识别方法,仅针对信号样本训练与测试,存在无法复现、维护难度大等问题。文中提出一种通信信号多特征提取与支持向量机算法融合相的识别控制优化算法。在对异常信号的比对过程中,根据移动通信的特性建立准确的信号模型,并使用支持向量机对大规模数据进行分类并实现识别控制。实验结果表明,与两种传统方法的相比,所提算法对信号有较强的识别能力,从而达到预期的目标。

In the process of mobile communication network optimization,it is difficult to identify and control the abnormal signals. However,the traditional abnormal signal identification method only aims at the training and testing of signal samples,and is unable to reproduce and difficult to maintain. A recognition control optimization algorithm based on the fusion of multifeature extraction of communication signals and support vector machine algorithm is proposed. In the process of comparing abnormal signals,the accurate signal model is established according to the characteristics of mobile communication,and the support vector machine is used to classify large-scale data and realize the recognition control. The experimental results show that,in comparison with the two traditional methods,the algorithm has a strong ability to recognize signals,so as to achieve the expected target.