Sharad Gangele

Curve Fitting Analysis of Internet Traffic Sharing Management in Computer Network under Cyber Crime

crime is growing fast day-by-day through the spreading of Internet around the world. Many under-developed countries are using dial-up-setup network where a call is connected only after a little extra effort. Some identified cyber crimes are writing abusive letters, giving threats to other, sending unwanted mails, hacking secret data, phishing attack on websites etc. The Internet traffic growth has some positive correlation with the cyber crime as justified by the Markov Chain Model based analysis of authors to examine the interrelationship between traffic sharing and blocking probability. This analysis has been extended for two-call basis also and model based relations are derived. These relations are complicated in terms of mathematical structure. This paper presents least square based curve analysis for Markov Chain model based relationship between traffic sharing and network blocking. The earlier suggested complicated relationship has been simplified in the form of straight line showing a good fitting. The Coefficient of determination has been computed showing the high value towards unity. It proves that simplified linear relationships perform well as a thumb rule for expressing the complex relationship between traffic sharing and network blocking probability. .

Least Square based Curve Fitting in Internet Access Traffic Sharing in Two Operator Environment

In Internet service market, at many places the competition occurs among internet service providers. The network suffers from congestion and users have priority to prefer that network having lowest congestion. During the repeated connecting attempt process the user has to maintain call-by-call effort. In this paper Markov chain model is used to establish a relationship between internet access traffic sharing and blocking probability of the network. The model based relationship has been simplified into the linear relationship. The accuracy of the fitting is examined through coefficient of determination. This process has eased up the multiple parameter based complicated relationship into a much simplified form. An average linear relationship is predicted for handy applications.