We seek to unearth which distribution best describes traffic behaviour to help solve the problem of congestion on campus from a mathematical point of view. Telecommunications traffic engineering (i.e. Tele -traffic engineering) is the application of traffic engineering theory to telecommunications. Tele-traffic engineers use their cognition of statistics including queuing theory, the nature of traffic, their practical fashion model, and their mensuration and computer simulation to make predictions and to program telecommunication networks such as a telephone network or the Internet. Internet traffic is a flow of data across the internet. This flow exhibits certain behaviour in accordance with a Probability distribution. Statistical analysis was performed to understand the characteristics of the traffic population. The empirical cumulative distribution process (CDF) together with essential statistical parameters were benchmarked. The goodness of Fit (GOF) test using the Anderson-Darling (AD) estimation method was applied to establish the best probability distribution model which describes the situation alongside with Probability plot. Deep-down opinion or analysis hints that if the period of time (or space) of grouping is increased well enough, the degree of data relationships will eventually become unimportant by scaling. The N-Flow traffic model was used to determine the burstiness of internet traffic in finite sessions. The Empirical and Theoretical CDF graph was used to determine the discrepancies between them. In the end, it was discovered that the data best fit normal distribution. From experimental and theoretical analysis, it is clear that the internet traffic behaviour in Ghana Technology University College is normally distributed.
Author(s) Details
Emmanuel Kwame Mensah
Department of Computer and Telecommunication Engineering,
Faculty of Engineering, Ghana Communication Technology, PMB 100 Accra-North,
Ghana.
Please see the book here:- https://doi.org/10.9734/bpi/rumcs/v7/2039G
No comments:
Post a Comment