Research And Implementation Of Key Technology For QoS Policy Management In Campus Network

With the rapid expansion of Internet-scale, how to manage the network, especially how to manage the policy of QoS and other complex network polic yhas became a serious problem.The traditional network management demands professional standards for networkadministrators, andtheopreration is complexand fallible. Policy-based network management is an effective way to solve this problem, but there are some problems, such as lack of common standards, implementation complexity, policy conflict detection, digestion problems and so on.The paper studies around the QoS policies managementof campus network, focusing on QoS policies conflict detection and digestion problems, system scalability, equipment transparency and other issues. The paper gives a scalable frameworkmodel of QoS policies management, anddesign a QoS policies conflict detection algorithm basedclass and QoS policies conflict digestion basedman-machine synergy based on it.For solving scalability, equipment transparency and other issues of NMS, the paper proposes a scalable QoS polic ies management framework model(EPM: Expanded Policy Managament). EPM framework modelprovids equipment transparencyusinga unifiedconfiguration and notation shield managed devices command differences; Upgrading the configuration model mapping libraries can easily expand the system's management capabilites for new management device, and to improve system scalability.For the behavior conflict of different types or the same type, the paper proposed a QoS policies conflict detection algorithm based on class(PBPC : Policy Based Policy Conflict Check). The PBPC algorithm based on checking flow classification,according to the policy to be configured flow classification categories, takes different flow classification detection methods separately,and then analysis the behaviors,classify the conflict on the base of the analysis' reason. Analysis shows that the PBPC algorithm compared to the classic HFalgorithm has larger detection range, achieving more easily and so on. Experimental results show that the algorithm can effectively detect the different types of behavior conflicts or the same type of behavior problems in the conflict.For the QoS policies d igestion problem about behavior conflictof different types or same type, the paper proposes a policy-based Conflict Resolution Mechanism of the man-machine collaboration. This mechanism is based on the man- machine and control collaboration,,carried out by the network administrator to participate in policy is automatically re-digestion or network planning based on the reason of man- machine.. The core of the system is based on the policy conflicts policy restructuring digestion method(PRBR: Policy Recombenation Based Resolved). PRBR algorithm to a new policy priority principle and the new conditions or new action principle of substitution principle for digestion, depending on the problem, the flow classification and QoS policies for popular reassembled into a new policy, configure the router control layer to achieve the intent of the network administrator. Algorithm effectiveness analysis showed that, PRBR algorithm can correctly digestion problems detected, shielding realize the differences router policy templates for different router manufacturers have versatility.