DYNAMIC NETWORK TRAFFIC PER HOST CONTROLLING SYSTEM USING NORMAL PREDICTION INTERVAL

Abstract

Bandwidth management is not a single technique or tool that can be used to control network traffic. Successful provision of managed network bandwidth is likely to involve the application of many tools encompassing a number of different techniques. In a network, every host behind a gateway may have limit of bandwidth usage set by the network administrator to control and manage the traffic so it is proper managed. Most of the network bandwidth control is done manually or using simple methodology and it may not be practical for some host which really requires a high amount of bandwidth rather than other host in the network behind the same gateway. Having the same limitation for each host behind a gateway without studying the previous bandwidth usage may not be a proper bandwidth controlling techniques to be implemented. Normal Prediction Interval (Normal PI) is one of the methods that can be used to predict what will be the future bandwidth usage for each host by studying the previous traffic used in t time limit. This paper will discuss about future allocation bandwidth with the previous data/sample captured in order to have automatic and dynamic bandwidth controlling system the usage of Normal PI can be used to satisfy the need.