大数据驱动和分析的舰船通信网络流量智能估计

Big data driven and analyzed ship communication network traffic intelligent estimation

作者单位:

1. 陕西师范大学 计算机科学学院，陕西 西安 710000;2. 青海交通职业技术学院，青海 西宁 8100161. School of Computer Science, Shaanxi Normal University, Xi′an 710000, China;2. Qinghai Communications Technical College, Xining 810016, China

大数据驱动;舰船通信;网络流量;智能估计;K-means聚类;神经网络big data-driven; ship communication; network traffic; intelligent estimation; K-means clustering; the neural network

研究大数据驱动和分析的舰船通信网络流量智能估计方法，提升网络流量智能估计效果。利用基于数据驱动的K-means聚类算法，提取具有标志性的有效通信网络流量数据；通过领域粗糙集算法，在有效通信网络流量数据内，提取流量数据时间序列特征；利用随机森林算法，剔除多余的时间序列特征，实现特征降维；在长短期记忆神经网络内，输出降维的时间序列特征，输出舰船通信网络流量智能估计结果。实验证明：该方法可合理提取有效舰船通信网络流量数据，有效提取并降维通信网络流量数据时间序列特征，可精准智能估计通信网络流量。The intelligent estimation method of ship communication network traffic driven and analyzed by big data is studied to improve the effect of intelligent estimation of network traffic. Based on the data driven K-means clustering algorithm, the iconic effective communication network traffic data is extracted from the historical ship communication network traffic data. Using domain rough set algorithm, time series features of traffic data are extracted from effective communication network traffic data. Random forest algorithm is used to eliminate redundant time series features and achieve feature dimension reduction. In the LMTN, the time series features of dimension reduction are outputed and the results of intelligent traffic estimation of ship communication network are outputed. Experiments show that this method can extract effective ship communication network traffic data reasonably. This method can effectively extract and reduce the time series features of communication network traffic data. This method can accurately and intelligently estimate communication network traffic.

参考文献：

[1] 赵龙文, 苌道方, 朱宗良, 等. 基于SARIMA-BP模型的港口船舶交通流量预测[J]. 中国航海, 2020, 43(1): 50–55+94

[2] 薛晗, 邵哲平, 潘家财, 等. 基于文化萤火虫算法-广义回归神经网络的船舶交通流量预测[J]. 上海交通大学学报, 2020, 54(4): 421–429

[3] 姚立霜, 刘丹, 裴作飞, 等. 基于EMD聚类的实时网络流量预测模型[J]. 计算机科学, 2020, 47(S2): 316–320

[4] 李校林, 吴腾. 基于PF-LSTM网络的高效网络流量预测方法[J]. 计算机应用研究, 2019, 36(12): 3833–3836

[5] 赵敏. 数据驱动下交互网络群智感知任务分配仿真[J]. 计算机仿真, 2021, 38(1): 476–480

[6] 阳杰, 白晓伟, 颜巍, 等. 基于分层数据搜索的数据驱动算法研究[J]. 固体力学学报, 2021, 42(3): 241–248

[7] 赵雅兰, 续欣莹, 任密蜂. 数据驱动框架下的非高斯批次过程最小熵性能评估算法[J]. 太原理工大学学报, 2019, 50(2): 251–254

[8] 麻文刚, 张亚东, 郭进. 基于LSTM与改进残差网络优化的异常流量检测方法[J]. 通信学报, 2021, 42(5): 23–40

[9] 李佳, 云晓春, 李书豪, 等. 基于混合结构深度神经网络的HTTP恶意流量检测方法[J]. 通信学报, 2019, 40(1): 24–33