Nowadays, the network data flow analysis always adopting simulation or benchmark patterns to estimate performance, and considering about current complexity or diversify applications of network, the actual results of simulation will be under suspicion. If the test system selects real network equipment to analyze data flow, the results will be more precise than virtual systems. The new characteristics of analysis data flow system will have the capability of simulating the data flow in real network and tested algorithms. For this reason, this research will base on NetFPGA ideas into data flow analysis. The process of IC design usually takes a long time and needs to plan and to analyze in advance. The IC design sent to the IC wafer foundries to make IC until the simulation of IC design is checked correctly. VHDL and Verilog hardware description language with FPGA (Field Programmable Gate Array) development tools to program the NetFPGA. It is helpful to decrease time of developing the NetFPGA and to build a research system. The traffic flow which the system generates will pass through access layer and core network equipment. It may cause bandwidth overload or traffic congestion situation between nodes if there is no QoS mechanism support. Therefore, this project designs and implements an adaptive QoS guarantee for network virtualized architecture to make better network management capabilities. This project designs a QoS Engine to provide the network service quality. The Engine includes three planes: data plane, control plane and management plane. Data plane function that monitoring the traffic and bandwidth module, it obtains the QoS relevant information. And it stores that convey to QoS agent module operating the algorithms for QoS management plane. Control plane mainly function is according to the decision of admission control module, managing the Traffic and bandwidth the network presently. QoS agent module in management plane will transport the QoS parameters of Plane-based to adaptive QoS strategy for  implementing policies. Meanwhile, cooperative strategy in QoS agent will pass QoS parameters to each other between the different NetFPGAs for ensuring the quality of service. Adaptive QoS strategy considers the policy rule to the QoS parameters. Finally, after computing the results sent to the admission control module. Then the admission control module can order the traffic control module and bandwidth control module for quality control. This research combines outcomes of the other subprojects to build the test environment, such as receiving the marked packets from switch/routing modules, and then monitor agent will collect the information tables from traffic and bandwidth monitoring modules, at the same time QoS agent sends the inspection information to the adaptive QoS strategy. Based on the computational results of QoS strategy processes the traffic and bandwidth control. The QoS Agent offers the whole QoS information to the other network nodes depend on the cooperative strategy. The admission control module first uses mandatory policy that is network management system offered, to make QoS strategic decisions. By means of the behavior-analysis results of application layer, the QoS Engine will have capability of the precise recognizing and data flow treatment. The truth network equipment of QoS engine will analysis and recognize the data flow, and verifying the analysis results, so that the research can collect variety of network information from different layers, and obtaining the analysis information. This project develops related algorithm and mechanism to gather statistics and analyze on huge flow of information. The important bandwidth information is recorded by reasonably simplifying the network environment for long-term analysis. The network management system will research and develop the possible policy according the information. The QoS engine can analyze the benefits of the variety of strategies, so that the manufacturer of network equipment can design new generation products by the supported information. Furthermore it can provide different network management strategies reference for network manager in creating the network management systems. QoS Mechanism for Network Virtulization Motivation Description