物联网自治区域内的网络路径规划建模研究

Iot Autonomous Region Network Path Planning Modeling Study

针对物联网的网络缺少统一的标准,当网络路径关键属性发生变化时,传统的路径规划模型进行自治区域网络路径选择方法,无法对差异网络下的路径综合权重进行归一化处理,只能选择较为优良的路径进行数据的转发和投递,导致路径发现效率低、自适应性弱的问题。提出建立动态权重算法的物联网自治区域内的网络路径规划模型。通过对物联网自治区域内的环境进行分析组建多路径规划模型,融合层次选择算法搜索不同状态下的路径规划关键属性,并以网络路径的带宽、时延、抖动、丢包率属性特征为路径选择的判决标准,构建属性的判断矩阵,利用特征根法计算每个判断属性的权重,进行归一化计算,在对各个候选路径的灰色关联系数进行网络路径加权排序,进而完成对网络路径规划优化。实验结果证明,改进方法具有较高的效率,而且适应性强。

In view of the lack of unified standard for the network in Interact of things （loT） , when the key attrib- utes of the network path change, the network path selection method for the autonomous region by using traditional path planning model can not carry on the normalized processing of the path comprehensive weight under the difference network, but can only choose a better path to carry on the data transmission and delivery, leading to the problems of low efficiency of path discovery and weak adaptivity. A network path planning model in the autonomous region of loT based on dynamic weighting algorithm was proposed in the paper. Through the environment analysis in the autono- mous region of IoT, a multi - path planning model was set up. The hierarchy selection algorithm was fused to search the key attributes of path planning in different states, and the attributive characters of bandwidth, delay, jitter and packet loss rate of the network path were taken as the decision criteria of the path selection, to construct the judgment matrix of attribute. The latent root method was used to calculate the weight of each judgment attributes and make nor- malization calculation. Then, the network path weighted ordering was obtained based on the grey correlation coeffi- cient of each candidate paths, so as to complete the network path planning optimization. The experimental resuhs show that the improved method has high efficiency and strong adaptability.