Dissertation > Excellent graduate degree dissertation topics show

Research on Traffic Measurement in High-speed IP Network

Author: ZhangZhen
Tutor: WangBinQiang
School: PLA Information Engineering University
Course: Communication and Information System
Keywords: Flow measurement Summary data structures The time hierarchical grouping sampling Bloom Filters Traffic management system
CLC: TN915.06
Type: Master's thesis
Year: 2009
Downloads: 77
Quote: 0
Read: Download Dissertation

Abstract

Flow measurement depth understanding of network inherent nature, the fundamental way to grasp the network operating conditions, QoS management, network optimization, as well as the strong support of traffic engineering. In recent years, with the rapid development of modern network technology, the Internet in the overall size and architecture occurred a significant change in the traditional flow measurement mechanism can not meet the current and future measurement needs, high-speed network traffic measurement is facing severe challenges. The paper combines the National 863 Program \scalability perspective, focusing on the high-speed backbone network flow measurement front-end processing algorithms and engineering technology. The main work of this paper is as follows: 1. Traditional flow measurement model lacks scalability shortcomings, one for high-speed network traffic measurement model. The model uses a packet \Based on this model, abstract flow measurement of two key technologies: traffic sampling measurement technology, and summary data structure. Use this model to a high-speed network traffic measurement, both to reduce the demand for resources, but also improve the speed of data flow analysis. 2 analyzes the shortcomings of static sampling mechanism based on traffic load adaptive hierarchical packet sampling algorithm. The algorithm uses the the predefined measurement error, layered packet sampling, adaptive prediction method of traffic load balance between resource consumption and accuracy. Experiments were carried out and based on the actual Internet data simulation results show that: the method having easy operability, controllability, but also self-adaptability and resources without losing accuracy. Restrictions for high-speed network flow measurement by the computing resources and storage resources, presents a summary of the data structure based on multidimensional counting bloom filter (Multi-dimensional Counting Bloom Filter, MDCBF). It will be a one-dimensional counting bloom filter structure (Counting Bloom filter, CBF), extended to support the business flow storage, query and statistical multidimensional structure. Through the implementation of \Simulation experiments were carried out based on computer-generated data and Internet data, the results show that: the method under the premise of ensuring the accuracy of having high space utilization and low computational complexity. Based on the accuracy of the flow measurement and real-time requirements, design a separate system structure before and after the end of the high-speed backbone network, real-time traffic management system, the structure eliminates the tight coupling of the front and rear end, enhance the flexibility of the system. Based on this structure, discussed in detail in the front-end system core algorithm FPGA (Field Programmable Gate Array: Field Programmable Gate Array) implementation, simulation results demonstrate the feasibility and effectiveness of the program.

Related Dissertations

CLC: > Industrial Technology > Radio electronics, telecommunications technology > Communicate > Communication network > General issues > Testing, operation
© 2012 www.DissertationTopic.Net Mobile