Design And Implementation Of Data Processing And Monitoring Subsystem Of Radar Simulation Platform

The lengthy development cycle of traditional radar has been unable to adapt to the rapid iteration of computing technology and hardware equipment.There will be a situation where the hardware equipment originally determined in the radar research and development process has fallen behind.At the same time,the internal structure of the radar is complex,and the development of radar is diversified,which leads to the obvious difference and low reusability of various components in the field of radar research and development Therefore,softwarebased radar has gradually become the mainstream radar developmental direction.Radar simulation platform is an integrated software platform of electronic warfare,radar and communication,designed and achieved by software radar technology.With the gradual upgrading of the radar simulation platform,the scale of simulation tasks and computing clusters are constantly increasing,which greatly increases the complexity of the entire system structure.In order to better grasp the operation of the platform,the health status of cluster equipment,the intermediate state of component algorithm calculation,the actual efficiency of component communication,and the operation record of the overall simulation need to be closely monitored.In accordance with the idea of a hierarchical and low-coupling design,the data monitoring related functions of the main task of radar simulation to form an independent system,which is very beneficial for the radar simulation platform to focus on simulation calculation and to avoid the drag of peripheral modules on the simulation task.Therefore,this article undertakes the above tasks and have designed and implemented the data processing and monitoring subsystems of the radar simulation platform.First,in view of the key problem that it is difficult to balance the efficient operation of simulation tasks and massive data storage and query in the cluster,this article selects Influx DB(an efficient data storage tool in the time series data field),Zoo Keeper(a mainstream distributed information coordination framework)and Flume(the most flexible data collection framework)as the technical foundation and proposes some solution.(1)Using the single-point Influx DB local storage method to achieve efficient data storage and query;(2)Metadata is extracted regularly and registered with the Zoo Keeper cluster to establish the data connection of each node;(3)Combining the custom Flume component with the Zoo Keeper monitoring mechanism to achieve data playback and simulation analysis function.Finally,a data scheme that supports data lightweight storage,asynchronous backup,and flexible invocation is achieved.It has been verified that the solution can successfully meet various functional requirements smoothly,and can compare and test network load in the real-time storage scenario of simulation data of the same scale.The solution proposed in this article only accounts for 1.9% of other general distributed data storage solutions,which is more suitable for radar.The running scene of the simulation platform.Then,aiming at the key problem that it is difficult to monitor complex data indicators in real time in complex scenarios and it is difficult to any node set request with low cost and high efficiency,this article puts forward some solutions:(1)This article customizes the dedicated Zoo Keeper component combined with the node online and offline response function implemented by relying on the Zoo Keeper cluster,designs a low-cost loosely coupled and efficient data monitoring request distribution method;(2)Using Flume framework’s efficient linkage capability,a real-time monitoring data transmission method is designed;(3)Referring to the idea of componentization,the multi-source data monitoring application supporting dynamic switch and convenient expansion is completed by using Flume custom source component and reflection mechanism.The dynamic and real-time monitoring scheme of multidimensional data index is realized comprehensively.It has been tested and verified that the solution can successfully satisfy various functional requirements.By simulating the production environment,the dynamic response delay of the solution is measured between20ms-200 ms,which can satisfy the daily requirements of users.Finally,on the basis of achieving the above scheme,the system function is further improved.Using the laboratory equipment,a cluster environment was built to complete the joint deployment of the platform and the subsystem.The operation verifies the availability of each functional module of the subsystem,and the results show that it can meet the various functions of the radar simulation platform.