公开(公告)号：US08630167B2

公开(公告)日：2014-01-14

申请号：US13185676

申请日：2011-07-19

申请人： Peter Ashwood Smith

发明人： Peter Ashwood Smith

IPC分类号： G06F11/00

CPC分类号： H04L41/0668 ,  H04L45/00 ,  H04L45/02 ,  H04L45/22 ,  H04L45/28 ,  H04L45/66

摘要： Link identifiers such as VIDs, selected from a defined range of values, are locally assigned by each node on a link state protocol controlled Ethernet network to each of its links or adjacencies. The link identifiers are assigned by the nodes such that each link or adjacency at the node is uniquely identified by a different link identifier. A link state protocol adjacency notification mechanism or other flooding mechanism is used to disseminate the locally assigned link identifiers to other nodes on the link state protocol controlled Ethernet network. The link identifiers are added by the nodes to their topology databases to enable detour routes to be locally calculated by the nodes on the network in a distributed manner. Upon occurrence of a failure, the link identifiers are used to source route traffic around the failure so that traffic may continue to traverse the link state protocol controlled Ethernet network.

摘要翻译： 从定义的值范围中选择的诸如VID的链路标识符由链路状态协议控制的以太网上的每个节点本地分配给其每个链路或邻接。 链路标识符由节点分配，使得节点处的每个链路或邻接被不同的链路标识符唯一地标识。 链路状态协议邻接通知机制或其他洪泛机制用于将本地分配的链路标识符传播到链路状态协议控制的以太网上的其他节点。 链路标识符被节点添加到它们的拓扑数据库中，以使得绕线路由由网络上的节点以分布式的方式在本地计算。 发生故障时，链路标识符用于引发故障周围的路由流量，以便流量可以继续遍历链路状态协议控制的以太网。

公开(公告)号：US08264949B2

公开(公告)日：2012-09-11

申请号：US11512671

申请日：2006-08-30

申请人： Ravi Ravindran ,  Peter Ashwood Smith ,  Hong Zhang

发明人： Ravi Ravindran ,  Peter Ashwood Smith ,  Hong Zhang

IPC分类号： G01R31/08

CPC分类号： H04L45/12 ,  H04L41/06 ,  H04L45/00 ,  H04L45/02 ,  H04L45/123 ,  H04L45/18 ,  H04L45/22 ,  H04L45/28

摘要： A weighting process may be used to select between alternate neighbors in a RAPID calculation to enable policy and/or traffic engineering considerations to affect the selection of an alternate path through the network. The information used to weight the neighbors may static administratively assigned weighting information or dynamic weighting information such as local statistical traffic condition information. The process may take into account the amount of traffic being handled by the current primary next hop for the destination, the available capacity of the available alternate neighbors, the ability of the alternate neighbors to handle the additional traffic, and other considerations. Weighting may occur after a set of available loop free alternate neighbors has been determined. Alternatively, weighting may occur before the RAPID calculation has been performed to cause the neighbors to be ordered prior to RAPID processing. This may enable RAPID calculation to stop without considering all available neighbors.

摘要翻译： 可以使用加权过程来在RAPID计算中的备选邻居之间进行选择，以使策略和/或流量工程注意事项能够影响通过网络的备用路径的选择。 用于加权邻居的信息可以静态地管理地分配加权信息或动态加权信息，例如本地统计交通状况信息。 该过程可以考虑由目的地的当前主要下一跳处理的业务量，可用备用邻居的可用容量，备用邻居处理附加业务的能力以及其他考虑。 在确定了一组可用的无环路交替邻居之后，可能会发生加权。 或者，可以在执行RAPID计算之前进行加权，以使得在RAPID处理之前对邻居进行排序。 这可能使RAPID计算停止，而不考虑所有可用的邻居。

公开(公告)号：US08228896B2

公开(公告)日：2012-07-24

申请号：US11525616

申请日：2006-09-22

申请人： Peter Ashwood Smith

发明人： Peter Ashwood Smith

IPC分类号： H04L12/28

CPC分类号： H04L47/10 ,  H04L63/123

摘要： At least part of the information in a packet's header may be verified so that a network element may rely on the information, e.g. to provide enhanced QoS and/or to identify the source. The information may be verified by a verifying agent (VA) or a transmitting network element. Where a VA is used, the VA may sign the header information issuing its private key of a private/public key pair. The signature may take the form of a digest, certificate, or other form. Once the header information has been signed, the client may transmit packets using headers that include the signed information and may include the VA's signature as an option in the header or in another area of the packet. The VA may also transmit packets directly. The signature may be decrypted by other NEs to extract the signed information, which can then be compared with the normal header information.

摘要翻译： 可以验证分组头部中的信息的至少一部分，使得网络元件可以依赖于信息，例如， 以提供增强的QoS和/或识别源。 信息可以由验证代理（VA）或传输网络元件验证。 在使用VA的情况下，VA可以签署发布私钥/公钥对的私钥的头信息。 签名可以采用摘要，证书或其他形式的形式。 一旦标题信息已经被签名，客户端可以使用包括签名信息的标题来发送分组，并且可以将该VA的签名包括在该报头中或该分组的另一个区域中作为选项。 VA也可以直接发送数据包。 签名可以由其他网元解密，以提取签名的信息，然后可以将其与正常的头信息进行比较。

公开(公告)号：US08223660B2

公开(公告)日：2012-07-17

申请号：US12148418

申请日：2008-04-18

申请人： David Allan ,  Malcolm Betts ,  Nigel Bragg ,  Peter Ashwood Smith ,  Paul Unbehagen

发明人： David Allan ,  Malcolm Betts ,  Nigel Bragg ,  Peter Ashwood Smith ,  Paul Unbehagen

IPC分类号： G01R31/08 ,  H04L12/28

CPC分类号： H04L45/16 ,  H04L41/06 ,  H04L45/02 ,  H04L45/28 ,  H04L45/32

摘要： Multicast capabilities of a link state protocol controlled network are used to accelerate the flooding advertisement of topology change notifications within portions of the network. This flooding mechanism may be particularly efficient in a network with a large number of two-connected nodes such as a ring network architecture. A control plane specific multicast group address is used when flooding topology change notifications, and a process such as reverse path forwarding check is used as an additional control on forwarding of the notification to prevent looping of control plane packets. Two-connected nodes insert a forwarding entry into their FIB to enable frames containing the control message to be forwarded via the data plane on to the downstream node so that propagation of the control message along a chain of two-connected nodes may occur at data plane speeds.

摘要翻译： 链路状态协议控制网络的组播功能用于加速网络部分内的拓扑变化通知的泛洪广告。 这种洪泛机制在具有大量双连接节点（例如环网架构）的网络中可能是特别有效的。 泛洪拓扑变化通知时使用控制平面特定的组播组地址，并且使用诸如反向路径转发检查之类的过程作为通知转发的附加控制，以防止控制平面分组的循环。 双连接节点将转发条目插入到其FIB中，以使包含控制消息的帧能够经由数据平面转发到下游节点，使得控制消息沿双连接节点链的传播可能发生在数据平面 速度。

公开(公告)号：US08184648B2

公开(公告)日：2012-05-22

申请号：US12487407

申请日：2009-06-18

申请人： David Allan ,  Peter Ashwood Smith

发明人： David Allan ,  Peter Ashwood Smith

IPC分类号： H04L12/28

CPC分类号： H04L12/437 ,  A61K31/337 ,  A61K33/00 ,  A61K47/12 ,  H04L41/0668 ,  H04L41/0672 ,  H04L45/28 ,  H04L45/66 ,  H04L49/351 ,  H04L49/70

摘要： A ring control protocol is used to establish a separate control plane for a plurality of physically dual homed devices to enable collections of dual homed devices to be represented by a single pair of addresses into the attached routed Ethernet network. The gateway devices analyze the passing ring control packets to create direct mappings for data packets to the routed Ethernet network. Thus, although the dual homed devices are treated as a ring from a control perspective, the data path is implemented to be direct so that data packets continue to flow directly from the dual homed devices to each of the attached gateway devices. In one embodiment, each of the gateway devices implements a virtual switch and advertises the MAC address of the virtual switch into the routed Ethernet network rather than the MAC addresses of each of the attached Ethernet Switch Units.

摘要翻译： 环控制协议用于为多个物理双重归属设备建立单独的控制平面，以使得双归属设备的集合可以由连接的路由以太网中的单对地址表示。 网关设备分析通过的环控制数据包，以创建到路由以太网的数据包的直接映射。 因此，虽然从控制角度将双重归属设备视为环，但是数据路径被实现为直接的，使得数据分组继续直接从双归属设备流向每个附接的网关设备。 在一个实施例中，每个网关设备实现虚拟交换机，并且将虚拟交换机的MAC地址通告给路由的以太网，而不是每个连接的以太网交换机单元的MAC地址。

公开(公告)号：US08149736B2

公开(公告)日：2012-04-03

申请号：US12728977

申请日：2010-03-22

申请人： Peter Ashwood Smith ,  Guoli Yin ,  William McCormick

发明人： Peter Ashwood Smith ,  Guoli Yin ,  William McCormick

IPC分类号： H04L12/28

CPC分类号： H04L45/742 ,  H04L45/02 ,  H04L45/44 ,  H04L45/54 ,  H04L45/745 ,  H04L67/104 ,  H04L67/1065

摘要： A distributed hash table is implemented to store routing information on a network. Node IDs exchanged in connection with implementation of a link state routing protocol are used as keys in the distributed hash table, and routes are stored at one or more nodes on the network. When a route is learned, the route is processed against the set of keys to determine which nodes should store the route. When a route is needed, the route is processed against the set of keys to determine which nodes should have the route information. The manner in which the route is processed against the set of keys is the same in both instances, so that the DHT may be used to store and retrieve route information on the network. The DHT may be implemented to store MAC addresses, IP addresses, MPLS labels, or other information of interest to enable routes to be stored and learned by network elements on the network.

摘要翻译： 实现分布式哈希表以在网络上存储路由信息。 在链路状态路由协议的实现中交换的节点ID被用作分布式哈希表中的密钥，并且路由被存储在网络上的一个或多个节点上。 当学习路由时，根据密钥集处理路由，以确定哪些节点应存储路由。 当需要路由时，根据密钥集处理路由，以确定哪些节点应该具有路由信息。 在两种情况下，路由处理方式相同，因此可以使用DHT来存储和检索网络上的路由信息​​。 可以实现DHT来存储MAC地址，IP地址，MPLS标签或感兴趣的其他信息，以使网络上的网络元素能够存储和学习路由。

公开(公告)号：US07911944B2

公开(公告)日：2011-03-22

申请号：US11964478

申请日：2007-12-26

申请人： Jerome Chiabaut ,  David Allan ,  Nigel Bragg ,  Peter Ashwood Smith

发明人： Jerome Chiabaut ,  David Allan ,  Nigel Bragg ,  Peter Ashwood Smith

IPC分类号： G01R31/08

CPC分类号： H04L45/12 ,  H04L45/00 ,  H04L45/22 ,  H04L45/24

摘要： A consistent tie-breaking decision between equal-cost shortest (lowest cost) paths is achieved by comparing an ordered set of node identifiers for each of a plurality of end-to-end paths. Alternatively, the same results can be achieved, on-the-fly, as a shortest path tree is constructed, by making a selection of an equal-cost path using the node identifiers of the diverging branches of the tree. Both variants allow a consistent selection to be made of equal-cost paths, regardless of where in the network the shortest paths are calculated. This ensures that traffic flow between any two nodes, in both the forward and reverse directions, will always follow the same path through the network.

摘要翻译： 通过比较多个端到端路径中的每一个的有序节点标识符集来实现等成本最短（最低成本）路径之间的一致的打破决定。 或者，通过使用树的分支分支的节点标识符选择等价路径，可以实时地实现与最短路径树相同的结果。 这两种变体允许对等成本路径进行一致的选择，而不管网络中哪些地方计算最短路径。 这确保任何两个节点之间在正向和反向方向上的业务流量将始终遵循通过网络的相同路径。

公开(公告)号：US20100271936A1

公开(公告)日：2010-10-28

申请号：US12431493

申请日：2009-04-28

申请人： David Allan ,  Peter Ashwood Smith

发明人： David Allan ,  Peter Ashwood Smith

IPC分类号： H04L12/24 ,  H04L12/28

CPC分类号： H04L12/462 ,  H04L45/02 ,  H04L45/021 ,  H04L45/18 ,  H04L45/22 ,  H04L45/28 ,  H04L45/66 ,  H04L69/40

摘要： A set of critical nodes or links is identified on the network and alternate forwarding state is pre-computed and disseminated within a node such that, upon failure of one of the critical nodes/links, a minimal trigger will cause the alternate forwarding state to be used to forward traffic on the routed Ethernet mesh network. In one embodiment rather than storing full tables of alternate forwarding state, only the changes to the forwarding state required by an identified failure is stored by the network nodes. Upon occurrence of a failure on the network, the identified failure is used to select the alternate forwarding state.

摘要翻译： 在网络上识别一组关键节点或链路，并且在节点内预先计算和传播备用转发状态，使得在关键节点/链路之一发生故障时，最小触发将导致备用转发状态为 用于在路由以太网网络上转发流量。 在一个实施例中，而不是存储替代转发状态的完整表，只有所识别的故障所需的转发状态的改变才被网络节点存储。 在网络发生故障时，使用识别的故障来选择备用转发状态。

公开(公告)号：US20100172270A1

公开(公告)日：2010-07-08

申请号：US12728977

申请日：2010-03-22

申请人： Peter Ashwood Smith ,  Guoli Yin ,  William McCormick

发明人： Peter Ashwood Smith ,  Guoli Yin ,  William McCormick

IPC分类号： H04L12/28

CPC分类号： H04L45/742 ,  H04L45/02 ,  H04L45/44 ,  H04L45/54 ,  H04L45/745 ,  H04L67/104 ,  H04L67/1065

摘要： A distributed hash table is implemented to store routing information on a network. Node IDs exchanged in connection with implementation of a link state routing protocol are used as keys in the distributed hash table, and routes are stored at one or more nodes on the network. When a route is learned, the route is processed against the set of keys to determine which nodes should store the route. When a route is needed, the route is processed against the set of keys to determine which nodes should have the route information. The manner in which the route is processed against the set of keys is the same in both instances, so that the DHT may be used to store and retrieve route information on the network. The DHT may be implemented to store MAC addresses, IP addresses, MPLS labels, or other information of interest to enable routes to be stored and learned by network elements on the network.

摘要翻译： 实现分布式哈希表以在网络上存储路由信息。 在链路状态路由协议的实现中交换的节点ID被用作分布式哈希表中的密钥，并且路由被存储在网络上的一个或多个节点上。 当学习路由时，根据密钥集处理路由，以确定哪些节点应存储路由。 当需要路由时，根据密钥集处理路由，以确定哪些节点应该具有路由信息。 在两种情况下，路由处理方式相同，因此可以使用DHT来存储和检索网络上的路由信息​​。 可以实现DHT来存储MAC地址，IP地址，MPLS标签或感兴趣的其他信息，以使网络上的网络元素能够存储和学习路由。

公开(公告)号：US07590067B2

公开(公告)日：2009-09-15

申请号：US10691517

申请日：2003-10-24

申请人： Hong Zhang ,  Peter Ashwood Smith ,  Guo Qiang Wang

发明人： Hong Zhang ,  Peter Ashwood Smith ,  Guo Qiang Wang

IPC分类号： G01R31/08

CPC分类号： H04L45/48 ,  H04L45/02 ,  H04L45/12

摘要： A method of computing allowable routes through a data network that includes a subnetwork that introduces a subnet intransitivity constraint on allowable paths through the network involves using an abstracted network map to construct a directed graph that compensates for the subset intransitivity constraint by removing all unallowable paths. The abstracted network map may be used to compute least cost allowable paths through the network. The directed graph represents the same network, and is constructed by representing bi-directional links between the subnetwork elements with two directed edges, and representing the subnetwork elements as paired ingress and egress nodes. A method for constructing the directed graph representing an abstracted map of a network that exhibits subset intransitivity is also described.

摘要翻译： 通过数据网络计算允许路由的方法，该数据网络包括通过网络在允许的路径上引入子网不可约束约束的子网络涉及使用抽象的网络映射来构造有向图，该有向图通过去除所有不允许的路径来补偿子集的传输约束。 抽象网络图可以用于计算通过网络的最小成本允许路径。 有向图表示相同的网络，并且通过在具有两个有向边缘的子网络元件之间表示双向链路并且将子网络元素表示为配对的入口和出口节点来构造。 还描述了一种用于构建表示呈现子集不透明性的网络的抽象地图的有向图的方法。