公开(公告)号：US06490249B1

公开(公告)日：2002-12-03

申请号：US09203104

申请日：1998-12-01

Applicant: Osama S. Aboul-Magd ,  Sameh A. Rabie

Inventor： Osama S. Aboul-Magd ,  Sameh A. Rabie

IPC: G01R3108

CPC classification number: H04L12/5602 ,  H04L2012/562 ,  H04L2012/5621 ,  H04L2012/5632 ,  H04L2012/5638 ,  H04Q11/0478

Abstract: An adaptive method and apparatus for regulating connection admission of traffic for networks, which includes a limiter that limits a volume of traffic admitted to a network through a connection admission control (CAC) procedure; and basing the connection admission control procedure at least in part on a pure measurement CAC that is solely based on measurements of actual traffic levels on the network. The connection admission control is based both on the pure measurement-based CAC and on a pure mathematical-based CAC that is solely dependent on user-supplied traffic parameters, thereby combining both types of the CAC to form a hybrid admission criterion. Where the network offers different services, managing link resources for CAC is effected by regulating the CAC procedure based on capacities assigned to service bandwidth pools that are defined for the different services offered by the network. The CAC procedure may take place on a virtual network (VN) environment.

Abstract translation: 一种用于调节网络业务连接准入的自适应方法和装置，其包括通过连接接纳控制（CAC）程序限制允许进入网络的业务量的限制器; 并基于连接准入控制程序至少部分地基于仅基于网络上实际流量水平的测量的纯测量CAC。 连接准入控制基于纯粹的基于测量的CAC和纯粹的基于数学的CAC，其仅依赖于用户提供的业务参数，从而将两种类型的CAC组合以形成混合准入标准。 在网络提供不同服务的情况下，通过根据为网络提供的不同服务定义的服务带宽池的容量调整CAC过程来管理CAC的链路资源。 CAC过程可能在虚拟网络（VN）环境中进行。

公开(公告)号：US07333438B1

公开(公告)日：2008-02-19

申请号：US10176060

申请日：2002-06-21

Applicant: Sameh A. Rabie ,  Tong Shan

Inventor： Sameh A. Rabie ,  Tong Shan

IPC: H04L12/26

CPC classification number: H04L45/10 ,  H04L41/0654 ,  H04L43/0811 ,  H04L45/22 ,  H04L45/28 ,  H04L45/302

Abstract: A recovery method for use in a connection-oriented network having nodes interconnected by links for establishing connections within the network, the method described enables the recovery of connections in the network in the event of a failure. The recovery method includes the following steps: (a) assigning connection re-routing priority to the connections in the network based on configured polices in response to the failure; and (b) creating release message bundles based on the assigned connection re-routing priority for use by the nodes in the network to recover connections. More specifically, a failure (e.g., link failure) could affect hundreds or thousands of connections passing through a node. To recover these connections, the node determines which connections should be recovered first and which connections can wait (or may never be recovered). The recovery method determines the priorities of re-routing connections after failure. After the re-routing priorities are determined for connections, the present invention provides mechanisms for bundling release messages and pacing the messages using hold-off timers to maximize the recovery of connections based on various network configurations and implementations.

Abstract translation: 一种用于面向连接的网络中的恢复方法，所述网络具有通过用于建立所述网络内的连接的链路互连的节点，所述方法使得能够在发生故障的情况下恢复所述网络中的连接。 恢复方法包括以下步骤：（a）响应于故障，基于配置的策略为网络中的连接分配连接重路由优先级; 和（b）基于分配的连接重新路由优先级创建释放消息束，以供网络中的节点使用以恢复连接。 更具体地，故障（例如，链路故障）可能影响通过节点的数百或数千个连接。 要恢复这些连接，节点确定首先应该恢复哪些连接，哪些连接可以等待（或可能永远不会被恢复）。 恢复方法确定故障后重新路由连接的优先级。 在为连接确定重路由优先级之后，本发明提供了用于使用保留定时器捆绑释放消息和起步消息的机制，以基于各种网络配置和实现来最大化连接的恢复。

公开(公告)号：US07092356B2

公开(公告)日：2006-08-15

申请号：US09970883

申请日：2001-10-05

Applicant: Sameh A. Rabie ,  Osama Aboul Magd

Inventor： Sameh A. Rabie ,  Osama Aboul Magd

IPC: H04L12/26

CPC classification number: H04L47/824 ,  H04L12/5602 ,  H04L45/10 ,  H04L45/24 ,  H04L45/302 ,  H04L45/308 ,  H04L47/15 ,  H04L47/70 ,  H04L47/786 ,  H04L47/805 ,  H04L47/822 ,  H04L47/829

Abstract: The invention relates generally to data communication networks and more particularly to a method of bandwidth management in a multiservice connection-oriented network which uses one or more overlooking factors and one or more overbooking models. The method allows an edge node which has received a connection request to accurately determine the bandwidth available on a given link in the network, by ensuring that different overlooking models and different overbooking factors are normalized at the edge node. The method of the present invention comprise: receiving at a selected one of the edge nodes updates from each of the core detailing a bandwidth management model, one or more overbooking factors and the link capacity associated with each of the core nodes; receiving at the selected one of the edge nodes a connection request; determining at the selected one of the edge nodes a preferred route through the network by accounting for a variation in overbooking factors or bandwidth management models between the selected one of said the nodes and each of the core nodes.