基于大数据统计的船舶通信网络流量估计数学模型设计

Design of mathematical model of ship communication network traffic estimation based on big data statistics

估计船舶通信网络流量,可以更好地了解网络性能,有助于优化网络配置,提高船舶通信网络的性能和可靠性,确保船舶航行过程中的通信畅通。为此,设计基于大数据统计的船舶通信网络流量估计数学模型,以提升网络流量估计效果。利用C-C算法计算船舶通信网络流量的延迟时间,通过G-P算法计算船舶通信网络流量的嵌入维数;依据延迟时间与嵌入维数,转换原始船舶通信网络流量时间序列数据,得到多维船舶通信网络流量时间序列数据;利用Map机制为各节点上的极限学习机分配数据子集,建立船舶通信网络流量估计数学模型;通过Reduce机制,汇总全部网络流量估计数学模型,得到最终的网络流量估计结果。实验证明,该模型可有效确定延时时间与嵌入维数,分别为5 min与6维;该模型可精准估计船舶通信网络流量。

Estimating the traffic of ship communication network can better understand the network performance,help to optimize the network configuration,improve the performance and reliability of ship communication network,and ensure the smooth communication during ship navigation.Therefore,a mathematical model of ship communication network traffic estimation based on big data statistics is designed to improve the effect of network traffic estimation.C-C algorithm is used to calculate the delay time of ship communication network traffic,and G-P algorithm is used to calculate the embedded dimension of ship communication network traffic.According to the delay time and embedded dimension,the original ship communication network traffic time series data is converted to obtain the multi-dimensional ship communication network traffic time series data.Map mechanism is used to assign data subset to the extreme learning machine on each node,and the mathematical model of ship communication network traffic estimation is established.The reduce mechanism is used to summarize all the mathematical models of network traffic estimation and get the final network traffic estimation result.Experimental results show that the model can effectively determine the delay time and embedding dimension,which are 5 min and 6D respectively.This model can accurately estimate the traffic of ship communication network.