Research On Technologies Of Software Monitoring System

Software monitor system plays an important role in large applications and key applications.However,current software monitor system should be improved on modulization and comprehensiveness and an unified software monitor standard has to be established as monitor information is processed and analyzed together when customization shows poor performance.This paper introduces the merits and drawbacks of ARM,analyzes and classifies common monitor tasks of applications,and establishes a software monitor standard.In our research,Compressed Sensing is applied to decrease performance requirement of monitor data transmission to improve the monitor software.This paper introduces Compressed Sensing technology and ways to lessen the transmission burden by improving the technology and data source.So we processed the unstable data source by wavelet transform and stable-increased one by Fourier transform to sparse domain for further processing by Compressed Sensing and finally to the system.As future status of the software is a great part in monitoring,CPU load is predicted by time series.This paper presents a multidimensional spatial voting prediction model to predict real-time CPU load accurately.We improved the real-time CPU load prediction accuracy by gray prediction model under the one-dimension prediction;we also applied voting mechanism to find a more appropriate classifier prediction model for predicting the CPU load in real time.Our experiments showed that multidimensional spatial voting prediction model led to better predictions than classic models with less mean relative error and mean square error.At the later part of the paper we tested the monitor system in real-time telecommunication marketing system to validate the standard and methods in this paper.