Abstract: Packets may be forwarded between realms employing private unregistered addresses without the use of network address translation and/or application level gateways. Nodes within privately addressed realms are identified by a combination of their locally significant address and a globally significant realm address. The globally significant addresses are reserved in all realms for use as realm identifiers or inter-realm routing. To send a packet to a node within a distinct private realm, the packet is given an inner IP header designating the locally significant destination IP address of the target node within the remote realm and an encapsulation IP header indicating the globally significant address advertised by the gateway of the target node's realm. The globally significant address is used for forwarding outside the realm of the destination node. Once the packet reaches the destination realm, the locally significant address is used for forwarding.

Abstract: A specialized preamble is provided to facilitate matrix channel estimation of a MIMO channel. In a particular implementation, a channel training preamble provided by the IEEE 802.11a standard is modified to facilitate MIMO channel estimation.

Abstract: Multiple Input Multiple Output (MIMO) technology in conjunction with the IEEE 802.11 standard enables simultaneous communication of data packets to or from multiple users in the same frequency. Spatial divisional multiple access (SDMA) is thus provided. In this way, system capacity can be increased to an extent that depends on available antenna resources and the multipath characteristics of the communication channel. Doubling or quadrupling of network throughput can be achieved.

Abstract: A system for detecting eye closure through optical observation of the eye includes a frame configured to be worn on a user, a light source, and at least one eyepiece connected to the frame. The eyepiece has a reflective surface on a side facing the user's eye when the frame is positioned on the user and the reflective surface of the eyepiece is configured for reflecting light emitted from the light source onto the eye. A sensor is connected to the frame and is operable to detect light reflected by the eye in a direction that is substantially parallel to the light incident on a retina of the eye to detect that the eye is in an open position.

Abstract: A crossbar switch is provided that can allocate output port bandwidth both among multiple classes of service and among input ports. In one implementation, there are two scheduling stages that may be implemented together. One scheduling stage can select traffic for output based on class of service whereas the other scheduling stage may perform a further selection based on destination output port. This allows greater flexibility in managing congestion and allocating limited available bandwidth.

Abstract: A multiprotocol encapsulation method and system for transporting different protocols over Ethernet are disclosed. The method includes receiving frames from an optical interface and identifying a protocol of the frame. The frames are multiprotocol and include storage network interface protocols. The frames are encapsulated into Ethernet protocol and transported over an Ethernet link.

Abstract: Local rerouting around a failed component link of a link bundle is provided by immediately substituting one or more other component links of the same link bundle. The substitution of component links is performed at the point of failure without signaling to other nodes. This minimizes signaling traffic particularly when large numbers of LSPs are impacted by a single component link failure. Also, since LSP repair can be accomplished very quickly, traffic disruption is minimized.

Abstract: Systems and methods for efficient storage of network management object identifiers are provided. Object identifiers such as SNMP MIB object identifiers are stored in a compressed form. In one embodiment, a single integer represents multiple elements of an object identifier where these elements are common to many identifiers. The redundancy inherent in a standard object identifier format is thereby greatly reduced. Large memory savings are provided in network management systems that employ many network management objects.

Abstract: A method for preventing loops in a ring topology of a computer network is disclosed. The method includes receiving a packet at a node on a shared packet ring and inserting a header containing an ID identifying the node at which the packet was received on the ring as an originating node. The packet is forwarded onto the ring and the ID is checked when received on a node. The packet is dropped if the ID identifies the node as the originating node.

Abstract: A Fast Reroute implementation where switchover time to backup tunnels upon failure of a protected network element is independent of a number of entries corresponding to forwarding equivalence classes forwarding over LSPs using that element. During normal operation of a packet forwarding device, adjacency information for a received packet is retrieved from a forwarding table based on a look-up of the packet's forwarding equivalence class. Upon failure of a link or node, the appropriate entries in this forwarding table are rewritten to implement the switchover to preconfigured backup tunnels. The switchover is effective even before the rewrite process has completed. Upon detection of the failure, forwarding processing shifts to a fix-up mode. During the fix-up mode, the look-up to the previously mentioned forwarding table is followed by a look-up to a backup tunnel adjacency table based on a pointer retrieved from the forwarding table.

Abstract: Techniques for multi-stage switching in network elements are provided. The switching fabric can include three stages and as connections are being made through the stages, the connections are selected to minimize the difference in the number of signals between different stages. Additionally, the switching fabric can include six identical chips where pairs of chips are utilized for each of the three stages.

Abstract: Systems and methods are provided for transferring physical layer control information from a central access point to individual subscriber units while maintaining transparency to higher layers. Adaptation of wireline MAC protocols to wireless applications is greatly facilitated. Subscriber unit power level may be controlled from the central access point.

Abstract: Techniques for monitoring nodes in a network are provided. A circular list of bins is maintained and each bin stores identifiers to zero or more nodes to be monitored. At periodic intervals, the nodes specified by a bin are processed such as by sending a polling message to the node. Additionally, responses from polling messages can also be processed at periodic intervals.

Abstract: Improved scalability and performance for label switched routers are provided. Large numbers of interfaces may be readily accommodated. In one implementation, label imposition is divided between ingress and egress linecards so that the interface supporting a given service can be the one to impose any needed corresponding label.

Abstract: A data compression system that dynamically adapts the degree of compression used in response to the priority of the data to be compressed is provided by virtue of one embodiment of the present invention. In one embodiment, only low priority data is compressed while high priority data is passed through to the network in uncompressed form. Also, the compression ratio can be varied depending on the priority of the data to allocate a larger portion of available bandwidth to higher priority traffic while imposing greater compression related delay on lower priority traffic.

Abstract: A method of calculating data traffic flow in a communications network, the communications network comprising a plurality of nodes including a plurality of source nodes and a plurality of destination nodes, the nodes being connected to one another by links, the method comprising: (a) obtaining data traffic flow measurements through said nodes and links as input data; (b) correcting said input data if inconsistencies are detected; (c) calculating upper and lower bounds of data traffic flow from a selected one of the plurality of source nodes to a selected one of the plurality of destination nodes using the corrected input data.

Abstract: The MAC layer of a wireless communication standard is modified to facilitate MIMO (multiple input multiple output) processing and thereby increase capacity and/or spectral efficiency. In one implementation, the IEEE 802.11a standard is modified in this way. A contention free period (CFP) is divided between MIMO communications and non-MIMO communications. An access point may maintain a list of nodes operating in MIMO mode and nodes operating in non-MIMO mode.

Abstract: An efficient distributed architecture for forwarding packets. The packet to be forwarded arrives in an ingress port, is processed by an ingress forwarding engine, transferred to an egress forwarding engine, and then transmitted via an egress port. An address-based lookup at the ingress forwarding engine identifies the correct egress forwarding engine and also identifies a translation index specifying the forwarding equivalence class (e.g., combination of address prefix and mask) of the packet. The egress forwarding engine then uses the translation index as a memory pointer to recover adjacency information with which to rewrite the packet header. The egress forwarding engine may maintain its adjacency information entirely locally without the need to share the information or propagate updates to ingress forwarding engines. This approach results in a scalable and highly efficient packet forwarding architecture.

Abstract: A method for forming an adaptive phased array transmission beam pattern at a base station without any knowledge of array geometry or mobile feedback is described. The approach is immune to the problems which plague methods which attempt to identify received angles of arrival from the mobile and map this information to an optimum transmit beam pattern. In addition, this approach does not suffer the capacity penalty and mobile handset complexity increase associated with mobile feedback. Estimates of the receive vector propagation channels are used to estimate transmit vector channel covariance matrices which form objectives and constraints in quadratic optimization problems leading to optimum beam former solutions for the single user case, and multiple user case. The new invention in capable of substantial frequency re-use capacity improvement in a multiple user cellular network.

Abstract: A method for providing voice communications at an optical amplifier site of a DWDM system is disclosed. The method generally comprises receiving an optical service channel at the amplifier site and converting an optical signal of the optical service channel to a digital electrical signal at the amplifier site. The optical service channel carries voice data and has a wavelength that is out of band with respect to wavelength bands of a multiplexed payload channel of the DWDM system. The method further includes sending the digital electrical signal to a router configured to support voice over IP.