Research On Edge Computing Communication Framework And Key Technologies For Heterogeneous Network Convergence

With the development of edge computing,network communication and other technologies,the scale of communication clusters is increasing,and the heterogeneous network structure is becoming increasingly complex.In the communication scenarios of heterogeneous network integration such as ocean exploration and emergency disaster relief,the demand for heterogeneous communication between different devices and homogeneous topology networking communication within similar device clusters is becoming increasingly urgent.In response to the above requirements,this paper studies and designs a edge computing communication framework for heterogeneous network convergence,including heterogeneous candidate network selection decision optimization,node communication path control in homogeneous clusters,scalable design of heterogeneous network protocols,etc.The main research contents are as follows:(1)Design the BiLSTM-IEW-AHP heterogeneous network vertical switching algorithm for the decision-making and switching problems of candidate networks in heterogeneous networks.Firstly,the BiLSTM algorithm is used to predict the trend of QoS parameter changes during the communication process based on time series.Next,calculate the overall impact of QoS parameters on communication quality in different business scenarios and optimize the weights using the i EW-AHP algorithm.Finally,compare the QoS parameter values with the QoS parameter weights on the network to determine whether network switching is necessary.By analyzing the future trends of communication networks and the requirements for network parameter weights in different business scenarios,network decision-making and switching are carried out to ensure the selection of the optimal network suitable for business scenarios and with good network quality during the communication process.(2)Design a routing algorithm based on high latency communication node perception to address the high latency communication problem caused by excessive node spacing in homogeneous clusters.Firstly,based on the communication status within the cluster,the BiLSTM algorithm is used to predict the attenuation trend of the signal with increasing distance,and the farthest distance between nodes that can establish efficient communication is calculated;Next,based on the elbow method and K-means algorithm,groups are divided according to the communication distance that can establish efficient communication,critical nodes are identified,and cost capacity maps are constructed for critical nodes;Finally,based on the network flow optimization algorithm with maximum cost and maximum flow constraint,the optimal matching link is calculated and mapped as the communication path between critical nodes.In the case of a large number of cluster nodes and a small number of high latency communication nodes,communication paths can be quickly optimized to reduce unstable communication within the cluster.Compared to the distance vector based topology control algorithm,the convergence time is reduced by 36% in the scenario where the total number of clusters is 100 and the number of high latency nodes is 2.(3)Aiming at the problems of multiple nodes and complex device types in heterogeneous network application scenarios,such as the difficulty of device information management and synchronization,and the complexity of protocol docking,based on Devops and Kub Eedge edge computing framework,a edge computing communication framework and system oriented to heterogeneous network extensible protocol fusion is designed and implemented;In the system experiment stage,the designed and implemented results were deployed and multiple sea trial experiments were conducted in the ocean exploration application scenario.Among them,the experimental content includes receiving and fusing data such as unmanned aerial vehicles and unmanned vessel clusters in ocean exploration missions,and integrating heterogeneous satellite communication networks to interact with land base stations;During the testing and acceptance stage,testing companies with CNAS and other qualifications conducted verification of indicators such as heterogeneous protocol docking,communication rate,communication link anomaly capture time,communication network switching time,and heterogeneous network switching strategy in accordance with GB/T 25000.51-2016.The results verified that the docking of system satellite equipment and various network protocols,communication link anomaly capture time,and network switching time were less than 1 second The communication rate can be stable at the maximum allowed rate of 2.2Mbps by the device,and the correct network selection can be made based on network conditions and business scenarios.