Research On Complex Network State Scanning And Analysis System Based On Path Optimization

The rapid development of network technology and the widespread application of network services have made the network structure become more complicated,various performance indicators of the complex networks are intricate.Data involved presents the trend of big data,accordingly the maintenance work has also shown explosive growth.All of these have put forward higher requirements for the operation quality and safety of complex networks.Real-time detection and data analysis of this type of network are important means to solve this problem.Since the marine monitoring and management network of North China sea branch of state oceanic administration(NBSOA)is a typical complex network,which has the characteristics of multiple parallel networks,complex network state,numerous nodes and real-time requirements.Therefore,the paper studies the methods of solving the path optimization problem on network route,real-time state scanning and online data analysis of complex networks based on the NBSOA management network topology.The main contents are as follows:Firstly,the development background of the marine monitoring management network are studied both local and abroad,and its development status and important significance in terms of network structure and monitoring means are discussed.Secondly,for complex networks with large scale,complex structure,and high real-time requirements,the path optimization problem based on network scanning and routing probing is studied.Considering the needs of network scale,load,and dynamic effectiveness of nodes,a multi-factor path optimization model is established,and an improved path search algorithm based on double layering and optimized Q-Learning isproposed to solve the problem of network scanning transmission efficiency.For the problem that the solution time of the shortest path of the network increases sharply with the increase of scale,a dual-layered strategy of dividing the network by combining k-core and modularity is proposed to reduce the network size reasonably and effectively.Thirdly,for the problem of real-time scanning of the dynamic environment caused by the change of the status of nodes,the reinforcement learning mechanism is introduced to dynamically sense the network,and the Q value is used to estimate the change of the path cost of the network and a path rewarding more is selected.For the problem of slow convergence of the algorithm,the adaptive learning factor and memory factor are added to optimize the update formula and improve the convergence speed.Fourthly,the proposed algorithm based on different power-law exponents(2 to 3)and different sizes of the complex network is compared with A* algorithm and Qrouting algorithm to verify the effectiveness of the algorithm.Finally,the NBSOA network system is designed and implemented based on the Java web development framework and the algorithm proposed.System is divided into two modules: network scanning service and network status monitoring and analysis sub-system,which provides 24 hours scanning and monitoring of the network nodes,sensing the network status dynamically,reducing network congestion,finding warnings.The system can also generate statistical charts and stacked charts in different time periods as needed,from hours,days,weeks,to months,quarters,years,which realizes visualization of network operation status.The system has been deployed in the monitoring center of the North China sea branch,and results are satisfied.