Researching On Algoirthms Of SIP Signaling Analysts In IMS Network

Researching on Algorithms of SIP Signaling Analysis in IMS NetworkAs the demand of network convergence is increasing more and more seriously in theinformation age, the IMS commercial pilots increase ceaselessly, and the development andapplication of IMS has already been a foregone conclusion.it will be a trend which cannotbe halted the development trend of IMS core protocol of SIP signaling analysis becomesthe new research hotspot. The irresistible trend of IMS make the analysis of its coreprotocol called SIP signaling becomes the new research hotspot. The SIP analytical studyof IMS is of great significance about the IMS network status, commercial status, userattributes, the comprehensive analysis and summary of prediction of market positioningand so on. Besides, the research has very important significance to case detection whichused the communication of information as an important supplementary.There will be massive messages captured from the IMS network in actual application,and quantities of resources will be wasted on the message analysis. Besides, the capturedmessages may be lost ineluctability as a result of influences from the signaling acquisitionenvironment. The missed messages may have a serious impact on the analysis results.The article is devoted to studying the algorithms that could improve the efficiency ofSIP analysis and the message recovery solutions in the case of information missing. Theresearch referred to improving the efficiency of SIP message analysis is of two asides, oneis on the basis of strings matching and another is on the basis of on the session processeswhich include numbers of messages.This paper proposed an efficient algorithm of SIP analysis based on the stringmatching after researching the application in SIP analysis of several common stringsmatching algorithm. Firstly, the algorithm combines the advantage of the BM algorithmapplying in short string matching and the QS algorithm applying in middle-and-long stringmatching of SIP analysis together. Considering the feature of SIP message and to meet theneed of actual analysis demand, the pattern strings are classified as short pattern stringsand middle-long pattern. it can avoid the low efficiency caused by parsing all the messageswithout classification very well. Secondly, two kind of interruption is proposed. Theanalysis process can be interrupted when the analysis demand is met. This can avoid theresource waste which could enhance the analysis efficiency. Experiments show that thisalgorithm is more efficient and the analysis time can be reduced by43%.Another research direction is about the analysis of session process which includes different kinds of SIP message. This article propose an algorithm called efficient SIPanalysis algorithm based on session process which make good use of the logicalrelationship between different messages in the same session, deriving the lamination ideaof the lamination general analysis algorithm which is embedded into the type analysisalgorithm, classifying the SIP messages in the same session into three hierarchies. Thispaper analyzes and determines the hierarchy of SIP messages in sessions that of highfrequency. The kernel of the algorithms which succeed in avoiding the repeatable analysisof massive redundancy messages and improve the efficiency effectively is to analyze themessages of different hierarchies with different degree. Experiments show that thisalgorithm is more efficient and the analysis time can be reduced by64%.This paper puts forward a new effective solution of message recovery applying to thephenomenon of message missing caused by the acquisition environment inevitably. Themessage recovery solution takes full advantage of the redundancy messages in the efficientSIP analysis algorithm based on session process and makes available. This solution issuitable to the analysis of session process and ensures efficiency with minimal impact ofthe analysis results. Experiments show that this algorithm refines the analysis algorithmsthis paper proposed and the message can be recovered by70%at the deficiency rate of0.1%.The algorithms this paper proposed are so well coordinated that the analysisefficiency could be improved from the inside out and the analysis algorithms are of wellperfection. The simulated results show that the proposed algorithms are of well analysisefficiency and message recovery efficiency when existing missing messages.