Forecast Method Of Network Traffic And Application In Network Management

It is more important to construct Internet in quantity in recent years in our country. With the increasing improvement of users' demand to businesses, quality and quantity become the same important. Only in case of this, the construct and development of Internet can become more scientific.In this paper, through forecast to traffic, we offer network managers the possible traffic information in future. In according to these information, network managers can take measure in advance to avoid descending of the network capability due to network congestion.In the network traffic collection step, we select the MRTG as the collecting tool through comparing and analyzing. We collect some traffic information of the key important interfaces in the laboratorial. After dealing with the traffic information, we store them into the database.In the network traffic forecast step, in according to the demand of the users, we compare some forecast methods and select the method based on the two-status markov chain and realize it using JAVA. Through analyzing the forecast result, we amend the method to improve the precision of the forecast.Finally, we apply this forecast method to the system of the network management, so that reducing the network faults and improving the network capability.