Based On Statistical Analysis Of Network Traffic Characteristics Test

Thanks to rapid development of modern communication and computer technology, the extent and complexity of computer network improve constantly. In the mean time, the variant e-businesses exponentially increase the total quantity of data transferred in the network. The first imperative is to improve networks management and to smooth networks operation because of traffic jam in network and poor network service occurred frequently. And it is necessary to grasp the basic characters and rules of network activities. In addition, understanding network activity characters is important foundation for network layout, network security, high performance protocol design and etc. Furthermore, the increase of network extent and data quantity should convert to the track of improving the quality of network operation and economic benefit. The most essential and most immediate way to achieve the conversion is to grasp the distribution of data flow and the utilization of resource in networks. Network flow measurement is the key point to achieve the conversion and to improve the network operation quality.This paper chooses relevant measurement methods pointedly to measure designator equipments and performance parameters from practical network environments. It analyzes data flow and network protocol according to the measured data and obtains the practical network flow characters. The flow statistical analysis method presented here makes immediate effect on statistics and analysis of network operation efficiency as well as diagnosis of network fault. It also provides significant reference to applying EXTREME serial switch.This paper introduces the characteristics of normal/abnormal network flow and presents abnormal flow management according to practical examples.