Estimating Buffer Size Using Positive Normal Distribution in Wireless Networks

Abstract

This thesis presents the use of positive normal distribution to estimate the minimum buffer size by estimating Cumulative distribution function of available bandwidth and streaming rate by using enhanced bandwidth estimation technique WBest for wireless networks. This has been implemented by using MATLAB environment. Evaluation and experimentation show that the simulation of BROS algorithm using positive normal distribution is more effective than using normal distribution for estimations in wireless networks. The algorithm which has been implemented in MATLAB is used by varying different parameters. Thus by using different parameters, namely, MTBBU, streaming rate, buffer size we have generated different results that seems to be effective in estimating buffer size which can further improve the performance of streaming multimedia applications in wireless networks.