The System Software Design Of Data Generator In Cascade Model

In the field of digital testing,due to the rapid growth of test dataset capacity,it is difficult for a single data generation device to take into account various test requirements.The integration of multiple devices to build ultra-large capacity data is a development trend.In the integration system,how to efficiently and quickly utilize the integrated equipment to generate the ultra-large capacity data set required for testing is the key to the improvement of system test efficiency.The method of cascade data generation system composed of multiple data generation devices proposed in this paper can effectively solve this problem.In order to achieve the system of the cascade data generation,this paper studies the system software from the aspects of large-capacity pattern data compression,storage,transmission and instrument control.The whole data generation system is mainly divided into two parts: computer-side control software and instrument-side software.The system software adopts the client/server mode,and the computer-side integration software is the client,and the instrument-side control software is the server.In the process of sending code data and control commands to each data generator in cascade mode,we should consider the efficiency of data transmission and storage of code data and fast parsing of control commands.Finally,the data generating device serving as the trigger source in the cascade mode outputs a synchronization signal to other devices,thereby realizing synchronous output of signals of the plurality of data generating devices.Combined with the above analysis,the main research contents of this paper are as follows:(1)Design of data compression processing.Since the computer-side control software is to achieve collaborative editing of the pattern data required by multiple data generating devices.In the process of design,a compression algorithm suitable for the pattern data is designed to reduce the capacity of data in the process of storage and network transmission.Through experiments,the compression rate of data is up to 99.52%,which effectively reduces the data capacity.(2)Design of store and acquirement for large-capacity data.For large-capacity code data access,ordinary I/O operations will show a slow processing speed.For this phenomenon,this paper uses memory file mapping to achieve high-efficiency store and acquirement of code data.Through experiments,the transfer speed of the memory file mapping mode is more than 10 times that of the normal I/O operation,which effectively improves the access efficiency of the pattern data.(3)Network transmission design of large-capacity data.Through UDP,the IP address of the device can be automatically obtained in the local area networks,and then transmitted through the multi-thread network to improve the transmission speed of the large-capacity code data.(4)Design of parameter management.The regular data editing is abstracted into a control commands.According to the user's operation,different processing functions are dynamically selected to ensure correct editing of the pattern data,and the interface operations are also synthesized into control commands to achieve control for different hardware modules.