Adaptive OLSR Based On Beacon Period Model For FANETs

Flying Ad Hoc Networks(FANETs)has become one of the hot spots of UAV cluster technology.As the speed range of the UAV cluster becomes larger and the task scenarios are diversified,the fixed routing parameters in traditional routing protocols are difficult to meet the network requirements in different scenarios.In response to this problem,this paper proposes an adaptive OLSR based on the beacon period model.The beacon period model trained by the Support Vector Machine adjusts the beacon message sending interval of the OLSR in real time and improves the network performance of the OLSR.The main research content includes the following three aspects:(1)TOPSIS based on entropy weight is proposed.When establishing the sample set of the beacon period model,it is necessary to use the optimal beacon message sending interval of the scene node as the sample label.In most cases,multiple network performance indicators that do not have a certain beacon message sending interval are superior to other beacon message sending interval,and the best beacon message sending interval cannot be selected by a simple comparison of performance indicators,so this paper uses this The model comprehensively evaluates the network performance of different beacon message sending cycles,and selects the best beacon message sending cycle of the scene node.(2)A beacon period model based on support vector machine is established.Existing research shows that the simulation data of the network simulator NS-3 can be used as a reference for real data.Therefore,this paper builds a sample set based on a large number of NS-3 simulation data.The sample characteristics are the network topology influencing factors that are only relevant to the information of the node and its neighbors.The sample label is the best beacon message sending interval of the node;the model is a multi-classification model,which utilizes One vs Rest of the support vector machine for offline training,so that the model can correctly classify the samples and obtain the best beacon message sending interval of the node.(3)An adaptive OLSR based on beacon period model is designed.The protocol improves the OLSR protocol.The output of the beacon period model established in this paper is used as the beacon message sending interval of the node.The input of the beacon period model is the network topology influencing factors obtained by real-time information processing of the node and its neighbors.When the real-time information of a neighbor node cannot be obtained by the node,the prediction value based on the mobile model is used to replace the real-time information of the neighbor node;at the same time,the protocol uses the mobile model to determine the link stability,and adds the link stability constraint when updating the route.Increased packet delivery rate.This paper verifies the effectiveness of the improved OLSR in different scenarios through three types of scene simulation experiments.Experimental results show that the improved OLSR has better network performance than the comparison protocols.