Abstract: A novel and useful apparatus for and method of interference reduction in a communications receiver. The invention first finds the channel estimate and the noise vector from the receive signal. An interference detector is then used to choose an appropriate compensation filter and select between adaptive and deterministic moving average models for noise prediction. After compensation filtering, a new channel estimate and noise vector is calculated. The new noise vector is used to determine noise whitening coefficients if the adaptive model is used. In the deterministic model case, the coefficients do not need to be calculated since they are already known, having been calculated a priori. The noise whitening coefficients are then used in an equalizer such as a Viterbi algorithm based equalizer.

Abstract: A method of distributing network resources fairly in an ATM network that is applicable in situations where a large number of clients establish connections to a single server. The method provides a way to more evenly distribute network resources when large number of clients desire connections to a popular server. A fair percentage value is assigned to a popular server and advertised in a PTSE. The fair percentage value is associated with a reachable ATM destination address and is flooded throughout the network. When a source node desires to connect to this server, the route is chosen in accordance with the fair percentage value. Rather than always selecting the best route, the source node chooses a route at random from among a percentage of the best routes. The switches can be configured with the fair percentage value either statically via a network manager or dynamically under program control.

Abstract: A system for assigning wavelengths in an optical switch to enable unicast, broadcast and multicast traffic to be sent through the switch. The optical switch enables data to be switched in the optical domain as opposed to conventional switching in the electrical domain. This permits a very high data capacity and data rate using well-known WDM technology. Tunable transmitters are configured to connect each input port to a different output port within the switch. During each cycle of the switch the transmitters are programmed to a particular wavelength, the wavelength determining the destination output port. Interface (I/F) cards issue requests to send data to the switch and receive acknowledgements indicating permission to send data to a particular destination port. A controller processes the requests received from the I/F card to generate a set of acknowledgements. A two-way round robin technique is used to select the destination ports for each I/F card based on the requests received.

Abstract: A communications transceiver constructed to minimize the additional external components required for interfacing to the communications medium. The invention includes a communications transceiver with a reduced number of filters comprising a host interface, a control module, a transmit portion and a receive portion. The host interface is coupled to a host computer and functions to exchange transmit, receive and control related data to and from the transceiver. A novel aspect of the invention includes using the same filter for both receiving and transmitting. The single filter plays a dual role, providing filtering function during both receive operation and transmit operation, thus serving to minimize the external components required, the cost of the device and its complexity. A switching device, such as an analog switch, connects the filter either to the output of the transmitter during transmit operation or to the output of the communications medium interface during receive operation.

Abstract: A method of calculating a complex node representation for logical nodes in a hierarchical peer group in a PNNI based ATM network. The summary information for default spokes, exceptions and bypasses is determined in a way that closely represents the underlying metrics associated with the peer group. A list of all border nodes in the peer group is generated and maintained and a matrix of tables is generated, one table for each metric per each class of service. The table is populated by the best value associated with the corresponding metric for a particular pair of border nodes within a particular class of service. Once the table is populated, each element is placed in a group or bin. The bins are generated by dividing the full range of the metric into a plurality of smaller subgroups or bins. The bin having a count greater than a threshold count is selected and 80% of the worst value for that group is used as the default spoke.

Abstract: A single voltage controlled oscillator (VCO) running at a closed loop controlled fundamental frequency whereby the harmonic products generated by the VCO are used for both transmission and reception. The use of a single VCO and associated supporting circuitry for both transmit and receive portions of the circuit serves to reduce the cost of the transceiver. The radio transceiver comprises a single synthesized oscillator having high harmonic content. The frequency conversions in the receiver make use of the harmonic frequency components of the same oscillator used in performing modulation during transmission.

Abstract: An arbiter utilizing a link list to arbitrate access between multiple data sources and a single destination. The arbiter is of the least recently used type whereby the data source that has not sent data for the longest time is given the highest priority. The arbiter provides an arbitration function in a simple manner and at high speeds. The arbiter utilizes a Non Empty Source Queue (NESQ) list that comprises only sources that are non empty, i.e., that have data ready to send. If a source queue chosen for data transmission still has data to send, it is placed at the end of the NESQ list. When a source queue becomes empty after the transmission of data, the source index is removed from the linked list. Conversely, when a source queue that was previously empty receives a new packet it is added to the end of the linked list.

Abstract: A method of synthesizing a CRC generator circuit for generating optimized code written in a high level hardware description language such as VHDL. The hardware description language code for implementing a CRC generator circuit has significantly less delay and utilizes fewer gates than prior art CRC generator circuit synthesis techniques. The method is operative to generate VHDL code which is then used to synthesize the circuit. The method iteratively generates the remainder equations for a CRC generator given the generator polynomial. During each iteration of the software, duplicate terms in each remainder equation are eliminated. The number of iterations is equal to the number of data bits the CRC generator is to process during each clock cycle. Once all the duplicate terms in the remainder equations are removed, the equations are sorted, rebuilt and translated into hardware description language code such as VHDL. The resulting optimized code may then be synthesized.

Abstract: A method of measuring the round trip time (RTT) of a virtual circuit (VC) utilizing the Traffic Management mechanism of ATM and adapting the CRM (also known as XRM) Traffic Management descriptor parameter accordingly. The CRM parameter is setup adaptively and individually for a particular VC in real time while traffic is flowing so as to achieve better performance of the network. The invention comprises CRM calculation means, a database for storage of RM cell related information, including but not limited to, RM cell assigned sequence numbers, transmission times, reception times, etc., means for filtering the transmission and reception times. The CRM calculator functions to track the transmission and reception of RM cells from the source end station using a unique sequence number assigned to each RM cell. Filters can optionally be used to smooth the data recorded that is derived from RM cell transmission events, e.g., calculated round trip time (RTT).

Abstract: An apparatus for and a method of reducing the packet length count processing of data packets in a network device. The apparatus and method determine the length and end of a packet at the ingress point of the network computing device, e.g., a switch or router, using a single counting function. After the counting function verifies the length of the data packet, a signal indicating the length of a data packet is generated. This packet length indicator signal is then propagated throughout the system along with the data packet itself. The packet length indicator signal is then used at all decision points that involve the data packet. The packet length indicator signal itself is generated so as to indicate the beginning of the start of the data packet in addition to the end of the data packet, thus indicating the length of the packet. This obviates the need for the length of the packet to be counted over and over again in the course of processing the data packet within the switch or router.

Abstract: An ATM switch which utilizes an all optical switching fabric to perform switching functions. The switch is based on fiber optics and Dense Wavelength Division Multiplexing (DWDM) technologies. The switch utilizes a centrally located optical switch fabric with multiple distributed interface cards. The switch module comprises an N X N passive star coupler which functions to pass the optical signal input present at any of its input ports to all of its output ports. Each output port has associated with it a unique wavelength. An optical filter set to the wavelength functions to block all wavelengths except for the wavelength associated with that particular output port. The ingress portion of each receive/transmit interface module comprises a tunable optical transmitter that is set to the wavelength corresponding to the desired destination output port.

Abstract: A method of searching utilizing a longest match based Radix Search Trie with variable length keys and having the ability to handle keys being prefixes of other keys. The method is based on the well known Patricia search trie algorithm. The address prefixes representing the keys for the tree are modified before being processed by the algorithm. A single byte is added to the beginning of each address prefix which is set equal to the length of the address prefix. The combined address length byte followed by the address prefix is used by the Patricia algorithm . When one address is the prefix of another, the length byte will make both addresses unique and distinct from each other. The conventional Patricia search can now be used since the keys have been made unique. A circular prefix list is maintained in descending numerical order comprising an entry for each distinct value of address length insure that the longest matching address is found.

Abstract: An apparatus for and a method of compressing the forwarding decision for a frame within a network device. A forwarding decision is compressed yielding a forwarding pointer that occupies far less memory space than the corresponding output port vector. The compressed forwarding pointers are stored in a forwarding table that is accessed using a hash function. A forwarding CAM is used to resolve conflicts in the hash table. The output port vectors are stored in an output port vector table that comprises a relatively small number of possible combinations of the port vector. A forwarding decision is made for each received frame by a forwarding processor in the device. The forwarding decision is represented by a compressed forwarding pointer that is stored in a table and associated with the received frame. At some later point in time, the frame is output to one or more destination ports in accordance with a corresponding output port vector.

Abstract: A method of transferring relatively short messages in an Asynchronous Transfer Mode network utilizing an emulated connectionless oriented technique. The method permits a short burst of data to be transmitted between users, utilizing a modified signaling scheme. The signaling scheme is modified so as to emulate a connectionless scheme whereby messages are transferred with no guarantee of delivery. The length of a message can be up to 47 bytes, i.e., the length of the data content of a single ATM cell. The method includes sending a setup message from the source user to the attached node requesting that a call of the fast type be set up. In response to this setup message, the attached node replies with a connect message indicating to the source user that it can immediately begin to transmit its short message.

Abstract: An apparatus for maintaining an ordered list which can store any type and number of data items. The ordered list can be implemented in hardware so as to enable fast, efficient maintenance of an ordered list of data. The data to be stored in the list is stored as a plurality of data cells each comprising any number of bits. The ordered list permits data to be read from, written to and deleted from the list. Data is written to the list using a push operation and data is deleted from the listing using a pop operation. A controller, processor or other source, provides the cell address information for each read, push and pop operation. The ordered list comprises a plurality of index units with each index unit comprising a 3 to 1 multiplexor whose output is connected to a data cell comprising a register or suitable data storage device. The mux selects data to its output from either (1) input cell data (2) the data cell in the previous index unit or (3) the data cell in the next index unit.

Abstract: A method of path restoration in an ATM network which attempts to minimize the loss of cells in the event a node or a link fails. The method is suitable for failures in point to point Switched Virtual Circuits (SVCs) in ATM networks and makes use of the bi-directional properties of point to point SVC calls. The method permits the construction of an ATM network which includes call path redundancy whereby if a node or link fails, another path meeting the requirements of the call (if one exists) will be used on an automatic basis. The alternative or redundant path is placed into service in such a way that the flow of data is substantially constant. The switching of the path occurs with almost no data loss and without the overhead time associated with Q.SAAL based detection and signaling.

Abstract: A method for creating and resolving split VLANs in ATM networks. The method utilizes a Network Management System (NMS) and a standard management protocol such as SNMP. The method monitors the network by continuously polling all the devices in the network. Status related information collected by the NMS is used to construct a physical and logical topology of the network. In response to the detection of one or more physical splits in the network, the NMS creates one or more corresponding logical splits in the network by splitting one or more VLANs. In the case of two split portions, the split VLAN can be created if a redundant LES and second LECS are located on the side of the split which does not include the primary LES. The second LECS is configured to use the address of the redundant LES in replying to requests to join an ELAN from the disconnected LECs.

Abstract: A method that provides a solution to the problem of overlapping branches in a point to multipoint call. The method is suitable for both Leaf Initiated Joins (LIJ) and existing point to multipoint call. The method has applications when adding a new leaf to an already existing point to multipoint connection and where the ROOT node has a corrupted point to multipoint view of the network. The method includes the overlapping node realizing that it is an overlapping node whereby it continues with the establishment of the downstream path in a conventional manner. The overlapping node then waits for the SETUP/ADD PARTY CONNECT to be received before continuing. Once the CONNECT is received, the overlapping node establishes an end point reference path to the point to multipoint root node over existing branches. The new call segments are established and the overlapping segments are then trimmed from the tree.

Abstract: An optical add drop module (OADM) to enable the construction of ring networks. WDM is utilized to construct a ring type network suitable for use in both WAN and LAN environments. The network is constructed having a ring topology with a plurality of nodes connected around the ring in daisy chain fashion. Direct connections between any two end users can be established to permit unicast connections. Multicast connections are also possible from a source to several destinations. Each unicast or multicast connection between nodes utilizes a separate specific wavelength. Devices are connected to the ring network via an optical add drop module. Wavelengths pass optically through the OADM device on nodes that are intermediary to the two end nodes on either side of the connection. Only the two end nodes transmit or pickup the optical signals on that particular wavelength corresponding to the connection.

Abstract: A method of selectively providing redundancy on a dynamic basis to end users that request it. The method establishes both a primary and a redundant connection for each connection requested to be redundant. The redundancy method of the present invention has application to PNNI networks and to any public network that uses signaling and routing in its operation, i.e., ISDN, frame relay, networks utilizing general signaling, voice networks, etc. A unique call reference number generated by the end user is carried unchanged from switch to switch along the path. Each switch along the routing path registers the call as a primary call in its database. Subsequently, once the primary connection is established successfully, the redundant connection is established. Each switch along the redundant path checks to see if the primary path already includes that switch.