Abstract: Backup peers in an asymmetrically organized computer network are organized into a "pool" of available devices that are activated as necessary. The network comprises a set of remote peers and a set of local peers directly associated with a central computational facility, the local peers facilitating connection between the remote peers and the central facility. During the capabilities exchange, the "primary" local peer to which the remote peer connects provides a list of backup peer devices to the remote peer. If the primary peer fails (or if the rate of message exchange falls below some predetermined threshold), the remote peer can use the backup information to access a backup peer from the pool. The invention allows each backup peer to be assigned to multiple remote peers, with the total number of backup peers determined by aggregate backup utilization (rather than simply assigning an individual backup peer to each primary peer regardless of whether such a high level of redundancy is justified).

Abstract: The present invention includes a method and apparatus for prioritizing and scheduling ATM cells that are competing for a given time slot for transmission onto a network. A segmentation and reassembly controller establishes a data structure within a memory. The preferable data structure includes a virtual connection time table and a virtual connection parameter table that use linked lists to index to one another. The segmentation and reassembly (SAR) controller includes a sorter for determining the priority of the ATM cells and selecting one for a given time slot and a microcode engine to compute next scheduled time slot and adjust the virtual connection parameter table for a given virtual connection. Coupled to the microcode engine is a random access memory for storing microcode programs so that the SAR may be programmed to prioritize and schedule in a different manner.

Abstract: A subsystem for communicating a private network signalling message over a packet network and bridges for communicating a Media Access Control (MAC) layer frame over an isochronous channel and for communicating an isochronous signalling frame over a nonisochronous network.

Abstract: An intermediate station produces, for an altered packet that includes multiple cells from a received packet and one or more altered cells that are substituted for the corresponding cells of the received packet, a .DELTA.-CRC remainder that is combined with the CRC remainder included in the received packet to produce a CRC remainder for the altered packet. The station first produces for each altered cell a bit pattern that has zeros in the bit positions in which the bits of the altered cell and the corresponding received cell match and ones in the bit position in which the two cells do not match. The station next encodes the bit patterns in accordance with the CRC code and produces corresponding .DELTA.-CRC values. It then manipulates the .DELTA.-CRC values to produce the .DELTA.-CRC remainder, and combines the .DELTA.-CRC remainder with the CRC remainder in the received packet. The result is the CRC remainder for the altered packet.

Abstract: A router is integrated onto a single silicon chip and includes an internal bus that couples multiple data receive and transmit channels to a central processing unit. The channels each have an external interface for connecting to different LAN or WAN networks. The serial channels are convertible into one or more time division multiplexed (TDM) channels. A time slot assigner (TSA) assembles and disassembles data packets transferred in TDM formats, such as ISDN. The serial channels are used for separately processing data packets in each TDM time slot. The TSA is programmable to operate with different TDM formats. A single direct memory access controller (DMAC) is coupled to each serial channel and an Ethernet channel and conducts data transfers on the internal router bus through a common port. The DMAC uses a novel bus protocol that provides selectable bandwidth allocation for each channel. The router architecture includes different interface circuitry which is also integrated onto the silicon chip.

Abstract: A port aggregation protocol (PAGP) dynamically aggregates redundant links between two neighboring devices in a computer network through the exchange of aggregation protocol data unit (AGPDU) frames between the two devices. Each AGPDU frame contains a unique identifier corresponding to the device sourcing the frame and a port number corresponding to the port through which the frame is forwarded. The exchange of AGPDU frames and the information contained therein allows the neighboring devices to identify those ports corresponding to the redundant links. Each device then dynamically aggregates its ports corresponding to the redundant links into a logical aggregation port (agport) which appears as a single, high-bandwidth port or interface to other processes executing on the device.

Abstract: A multicast engine of a shared-memory switching fabric circuit increases the replication rate of data elements destined for multicast connections within a network switch by manipulating address information relating to those elements. The multicast engine cooperates with other components of the switching fabric circuit to minimize the total buffer requirements of the switch by storing only a single copy of each multicast data element in a location of shared memory. Specifically, the engine has a pipelined architecture that provides a multicasting capability for the switching fabric circuit by replicating only an address pointer to that memory location for each destination of the multicast connection.

Abstract: An arrangement facilitates rendering of a forwarding decision for a packet transferred between high-speed switches of a network using a path identifier (PTID) encapsulated within the packet. The PTID is a tag associated with a destination address of the packet and comprises a value that uniquely identifies a corresponding entry of a forwarding table of the switch. Rather than requiring a full compare of the destination address with the contents of the forwarding table, the PTID is used as a direct index address into the table that identifies the corresponding entry in a fast and efficient manner.

Abstract: A method and system for providing a virtual interface between a router and a network, in which the router is not connected to the network using a local interface. A method and system in which a router may be dynamically decoupled from a first network and coupled to a second network, without losing state information associated with the coupling to either network. The router comprises a virtual interface to the network, at which state information for the physical interface to the network is recorded, and a binding between the virtual interface and a physical interface, the latter of which is coupled to the network. Dynamic binding of the virtual interface to the physical interface comprises an authentication step. A method and system in which a router can be dynamically coupled to one of a plurality of local networks of differing types, such as one local network using an ethernet technique and one local network using a token ring technique.

Abstract: An IP address is automatically discovered by a network endpoint, such as a PC or router. The endpoint listens for a broadcast network packet or promiscuously listens for a unicast network packet sent from a web browser from a host system. The network packet includes the IP address for the host system and a preselected IP domain name. The IP domain name is used to initiate the address discovery in the endpoint. The IP address from the host system is used by the endpoint as a seed for generating a proposed IP network address. The endpoint then uses an address resolution protocol (ARP) to determine whether the proposed IP address is currently assigned to any other device in the network. If no device in the network responds to the ARP request, the proposed IP address constitutes a unique address on a network segment. Because the proposed address is not used by any other device in the subnetwork, it is self-assigned to the endpoint.

Abstract: A communication server includes a modem pool (652) that comprises a plurality of XDSL modems (660). The communication server also includes a plurality of line interface modules (650). Each line interface module (650) has a plurality of inputs each coupled to a twisted pair data line and has a plurality of outputs each coupled to one of the XDSL modems (660). Each line interface module (650) operates to detect a need for data service on a requesting twisted pair data line and to couple the requesting twisted pair data line to an output associated with a selected XDSL modem (660). A controller (653) is coupled to the XDSL modems (660) and to the plurality of line interface modules (650). The controller (653) operates to select one of the XDSL modems (660) in response to a detected need for data service.

Abstract: A method of reducing a number of bits required to transmit a digital signal using a PPP protocol is disclosed. The PPP protocol uses a predetermined byte for signalling. When that byte occurs, the following byte indicates signalling or alternatively, indicates a data byte having a value of the signalling byte. A value in the signalling byte is used to indicate a data byte, and other values are encoded to represent two or more signalling bytes within a predetermined spacing. Due to the look-ahead requirements, a time lag is introduced.

Abstract: A method for communicating packetized data over a communications link having a predetermined format including a framing using a channel bank having a time division multiplexing addressing scheme is disclosed in which a plurality of channel units may generate packetized data and the communications link may be allocated to a channel unit that is currently sending data over the communications link. The next sender of packetized data over the communications link may be determined by a system distributed over the plurality of channel units so that each channel unit independently determines whether it is the next sender. The communications link may be allocated to the channel unit that is the next sender of packetized data after the current sender of packetized data so that the time division multiplexing addressing scheme of the channel bank is ignored and the channel bank communicates packetized data. An apparatus for communicating packetized data over a channel bank is also disclosed.

Abstract: The invention provides a method and system for auto-sensing LMI protocols in frame relay networks. When a router is first coupled to a frame relay network, it automatically configures the local management interface (LMI) to use one of a selected set of possible LMI protocols, by generating a set of protocol requests for a plurality of protocols, and by thereafter simultaneously listening for protocol responses from the configuration server. Multiple valid responses from the configuration server are assigned priority in response to which valid response is last to arrive.

Abstract: A pluggable port adapter is used for connecting PCI devices to a host system through a PCI local bus while also adding functionality to the host system. The port adapter communicates with the host system through a port adapter/host interface that includes the PCI local bus and an auxiliary bus. The auxiliary bus is used for controlling the additional circuitry on the port adapter. A PROM on the adapter card is used for identifying the port adapter type, serial number and hardware revision. The auxiliary bus is used for conducting JTAG testing and is used by the host system to program logic devices on the port adapter. The logic devices can be reprogrammed in the field by the host system to repair bugs and to enhance performance and/or functionality. A power control circuit on the port adapter is controlled by the auxiliary bus for conducting hot swap operations.

Abstract: A special memory overlay circuit uses a first DRAM buffer memory in combination with a second faster SRAM buffer memory to reduce the time required to translate information into different network protocols. Packet data is stored in the DRAM buffer memory and packet headers requiring manipulation are stored in the SRAM buffer memory. Because the SRAM has a faster data access time than the DRAM buffer memory, a processor can reformat the packet header into different network protocols in a shorter amount of time. Packet headers also use a relatively small amount of memory compared to remaining packet data. Since the SRAM buffer memory is only used for storing packet headers, relatively little additional cost is required to utilize the faster SRAM memory while substantially increasing network performance.

Abstract: In a data transmission system having a first and a second data transmission device and a transmission line that is not terminated, a method of minimizing signal reflection along the transmission line is described that includes the step of connecting the first and second data transmission devices to the transmission line such that the distance between the first and second data transmission devices is substantially proportional to a wavelength of a predetermined frequency. Data to be transmitted from the first data transmission device to the second data transmission device via the transmission line are encoded in the first data transmission device such that energy of the encoded data is substantially concentrated around the predetermined frequency to minimize the signal reflection without terminating the transmission line. A data transmission system with minimized signal reflection is also described.

Abstract: A switch in an asynchronous transfer mode system utilizes an early packet discard (EPD) scheme and discards all but the end-of-packet cell of packets that are expected to prevent other partially transmitted packets from being transmitted through the switch. The switch also utilizes an integrated tail packet discard (I-TPD) scheme and, once a cell of a packet has been discarded for any reason, discards all of the remaining cells of the packet except the end-of-packet cell. Each of the EPD and the I-TPD schemes retain the last cell of a packet in order to maintain packet boundaries. Further, they each set the loss-priority of the end-of-packet cell to high, to increase the likelihood that the cell will be transmitted through the remaining switches in the route to the station to which the packet is directed. When the end-of-packet cell is received, the switch retains the cell unless its maximum queue limit is exceeded, and resets any associated early or tail packet discard flag.

Abstract: An encapsulation mechanism efficiently transports packets between ports of different switches in a network on the basis of, inter alia, virtual local area network (VLAN) associations among those ports. The switches are preferably interconnected by a novel interswitch link (ISL) mechanism that appends ISL destination and source information, along with ISL error detection information, to VLAN-modified packets. The ISL mechanism keeps the VLAN associations of the packets intact during transfer between the switches in accordance with a high-performance switching bus architecture.

Abstract: An address translation mechanism quickly and efficiently renders forwarding decisions for data flames transported among ports of a high-performance switch on the basis of, inter alia, virtual local area network (VLAN) associations among the ports. The translation mechanism comprises a plurality of forwarding tables, each of which contains entries having unique index values that translate to selection signals for ports destined to received the data frames. Each port is associated with a unique index value and a VLAN identifier to facilitate multicast data transfers within the switch at accelerated speeds and addressing capabilities.