EP0861469A1 - Technologie pour reseau a grande largeur de bande - Google Patents

Technologie pour reseau a grande largeur de bande

Info

Publication number
EP0861469A1
EP0861469A1 EP96940417A EP96940417A EP0861469A1 EP 0861469 A1 EP0861469 A1 EP 0861469A1 EP 96940417 A EP96940417 A EP 96940417A EP 96940417 A EP96940417 A EP 96940417A EP 0861469 A1 EP0861469 A1 EP 0861469A1
Authority
EP
European Patent Office
Prior art keywords
data packets
broadcast
input
network
ouφut
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Withdrawn
Application number
EP96940417A
Other languages
German (de)
English (en)
Inventor
Roger E. Billings
Michael O. Schmidt
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Wideband Corp
Original Assignee
Wideband Corp
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Priority claimed from US08/687,777 external-priority patent/US6115747A/en
Priority claimed from US08/690,676 external-priority patent/US5793981A/en
Priority claimed from US08/688,799 external-priority patent/US6061730A/en
Priority claimed from US08/690,673 external-priority patent/US5761433A/en
Priority claimed from US08/690,735 external-priority patent/US6076117A/en
Application filed by Wideband Corp filed Critical Wideband Corp
Publication of EP0861469A1 publication Critical patent/EP0861469A1/fr
Withdrawn legal-status Critical Current

Links

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L12/00—Data switching networks
    • H04L12/28—Data switching networks characterised by path configuration, e.g. LAN [Local Area Networks] or WAN [Wide Area Networks]
    • H04L12/44—Star or tree networks

Definitions

  • the present invention relates to computer networks and, more particularly, to computer networks in which data is communicated among a plurality of computers.
  • Computer networking involves moving data from one point to another.
  • Computer networking or "interoperation” usually begins with computers in the same office or the same building connected in a local network.
  • the term "local area network” or LAN describes a group of computers typically connected by no more than 1,000 feet of cable which interoperate and allow people to share resources.
  • Databases are becoming larger and more sophisticated and are being accessed by greater numbers of users. Further, many industries have been migrating toward applications involving high-resolution color graphics, which applications require high bandwidth to handle the large amount of data involved with this type of graphics.
  • Ethernet adapters with data rates of up to 100 megabits per second (Mbps) are now commercially available and are quickly becoming affordable.
  • Another approach known as “switching hub technology,” dedicates a portion of the LAN to a single or small group of users.
  • Many Token Ring LANs have also risen to the challenge to find ways to increase their performance and are operating at 16 Mbps.
  • ATM Asynchronous Transfer Mode
  • Network interface cards sometimes called LAN adapters, function as an interface between the computer and the network cabling, linking the computer to the network cable system.
  • the card controls the flow of data between the computer's internal data bus and the serial stream of data on the network cable.
  • Some computers are provided with a network interface card on the motherboard of the computer, but the cards are usually added to the computer's expansion bus.
  • Network interface cards also change the form of the computer generated data from a wide parallel stream, e.g., 8 bits at a time, to a narrow stream moving 1 bit at a time in and out of the network port.
  • Conventional network interface cards buffer data because the computer is typically much faster than the network.
  • Many network interface cards have a processor specially designed for on-board processing power. These processors are augmented by 8K to 64K of RAM and by specialized transceivers that handle the electrical interface with the cable.
  • Network interface cards generate the electrical signals that travel over the network cable. Each network interface card follows specific rules controlling access to the cable. Network interface cards for Ethernet and Token Ring both use the same basic system of electrical signaling of the cable. A technique called Manchester encoding provides a way to transmit Os and ls using direct current voltage pulses that range from - 15 to + 15 volts. The network interface cards translate each eight changes in the voltage level as a character in the ASCII data alphabet.
  • each data packet issued by each computer in the network is received by all the other computers in the network.
  • a computer monitors the network to determine when the network is idle, at which time the computer may issue a packet. If the network is busy, the computer must wait. If two or more computers determine that the network is idle and respectively issue packets at the same time, a collision occurs.
  • busy Ethernet networks collisions occur frequently.
  • the interface cards share the common cable by listening before they transmit and transmitting only during a break in the traffic when the channel is quiet. This technique is called carrier-sense multiple access with collision detection (CSMA/CD). With collision detection, if two stations begin to transmit at the same time, they detect the collision, stop, and retry after a sufficient time interval.
  • CSMA/CD carrier-sense multiple access with collision detection
  • Network interface cards for Token Rings use a complex media-access control scheme called "token passing." Whereas Ethernet cards contend for access to the cable, Token Ring cards must have permission to transmit into a cable system that forms a complete electrical loop or ring. Under this technique, the active cards negotiate, using their built-in serial number, to determine a master interface card. The master initiates a special message called a free token. When an interface card with data to send receives a free token, it changes the free token into a message and sends it to the next station up the ring for relay. After the addressed interface card receives the message and the message returns to the originating interface card, that card initiates a new free token and the process begins again.
  • a computer must wait for the marker before issuing a packet. Further, if a packet is addressed incorrectly or is otherwise undeliverable, the packet will endlessly circulate in the token ring. Therefore, special means must be provided to keep the ring clear.
  • ARCNET network interface cards use a media-access control scheme.
  • a designated “master card” maintains a table of all active network interface cards and polls each network interface card in turn, giving permission to transmit.
  • ARCNET uses a star cabling pattern with passive and active hubs that can extend the cabling farther than Ethernet or Token Ring configurations.
  • Networked devices share the cabling using an orderly polling scheme.
  • the installer sets switches on each network interface card, which gives the card a specific number.
  • the lowest numbered active card becomes a master controller. It sends a message to each adapter in sequence, giving it permission to transmit any data it holds.
  • the standard ARCNET signaling speed of 2.5 Mbps limits the maximum throughput, although faster speeds have been recently developed.
  • An ATM network is configured much like a telephone network. An issued packet travels from the issuing computer to the receiving computer on a communication line that other computers in the network are not currently sharing. Application software often needs to be modified to be compatible with an ATM network.
  • a computer network includes a plurality of computers each having an interface, a plurality of data channels each connected to a data output of the interface of one of the computers, a broadcast channel connected to a broadcast input of the interface of each of the computers, and a network hub connected to each of the data channels and the broadcast channel.
  • Each of the computers has a unique address and is capable of generating addressed data packets intended for delivery to another computer on the network.
  • the network hub includes a plurality of ports each of which is connected to a respective one of the data channels for receiving addressed data packets generated by one of the computers, and a broadcast output connected to the broadcast channel. The network hub sequentially merges addressed data packets received at each of the ports into a stream of data packets.
  • the network hub includes a plurality of storage devices and a merging processor.
  • Each of the storage devices is connected to a respective one of the ports, and receives and temporarily stores the addressed data packets generated by the computer connected to the respective port.
  • the processor is connected to each of the storage devices and sequentially merges addressed data packets temporarily stored by each of the storage devices into the stream of data packets.
  • An internal bus is connected to each of the storage devices and to the broadcast output for delivering the stream of data packets to the broadcast output.
  • the computer network includes a plurality of network hubs, each of which includes an expansion port.
  • the expansion port is connected to the internal bus so that the expansion port receives the same stream of data packets that are delivered to the broadcast output of the hub.
  • the expansion port of one of the network hubs is connected to one of the ports of another network hub. Therefore, the stream of data packets from one network hub may be merged with the data packets received at the ports of another network hub, thereby forming a combined stream of data packets which is broadcast to the computers of the network.
  • the network hub may communicate with the interface of each of the computers to indicate whether the computer may transmit data packets to the network hub for merging. Specifically, when one of the storage devices of the network hub is filled or nearly filled to capacity with temporarily stored data packets, the network hub may send a status signal to the interface corresponding thereto, indicating that the storage device is near capacity. Accordingly, each of the interfaces includes a local storage device for temporarily storing addressed data packets generated by the computer upon receiving the status signal from the network hub, until the merging processor has emptied the storage device.
  • FIG. 1 is a block diagram of an exemplary embodiment of a computer network implemented according to networking principles of the present invention
  • FIG. 2 is a block diagram of a computer network illustrating networking principles of the present invention, particularly illustrating a chain of stations connected by a sequential reception channel and a separate delivery channel;
  • FIG. 3 is a block diagram of an exemplary embodiment of a network hub configured in accordance with networking principles of the present invention
  • FIG. 4 is a block diagram of an exemplary embodiment of a network interface card configured in accordance with networking principles of the present invention
  • FIG. 5A is a block diagram of an exemplary method of merging data packets implemented in accordance with the networking principles of the invention.
  • FIG. 5B is a block diagram of another exemplary method of merging data packets implemented in accordance with the networking principles of the invention.
  • FIG. 6 is a block diagram of an exemplary embodiment of a network having mirrored servers implemented in accordance with networking principles of the present invention;
  • FIG. 7 is a block diagram of an exemplary embodiment of a network hub for use in a mirrored-server network configured in accordance with the invention.
  • FIG. 8 is a block diagram of an exemplary embodiment of a wide area network implemented in accordance with networking principles of the present invention.
  • FIG. 9 is a schematic view of an exemplary embodiment of a computer network implement in accordance with the present invention, particularly illustrating a file-server network configuration
  • FIG. 10 is a schematic view of an exemplary embodiment of a network interface card implemented in accordance with the invention.
  • FIG. 11 is a schematic view of an exemplary embodiment of a computer network implemented in accordance with the invention, particularly showing a dual file-server network configuration;
  • FIG. 12 is a schematic view of an exemplary embodiment of a network hub implemented in accordance with the present invention.
  • FIG. 13 is a schematic view of an exemplary embodiment of a computer network implemented in accordance with the invention, particularly showing a wide area network configuration
  • FIG. 14 is a schematic view of an exemplary embodiment of a computer network implemented in accordance with the present invention, particularly illustrating a wide area network configuration with access to the Internet;
  • FIG. 15 is a schematic view of an exemplary embodiment of a computer network implemented in accordance with the present invention, particularly showing a wide area network configuration with access to a mail server;
  • FIG. 16 is a schematic view of an exemplary embodiment of a computer network implemented in accordance with the present invention, particularly showing a mirror-server configuration
  • FIG. 17 is a schematic view of an exemplary embodiment of a computer network implemented in accordance with the present invention, particularly illustrating a video-server network configuration
  • FIG. 18 is a block diagram of another exemplary embodiment of a network hub illustrating packet-merging principles of the present invention
  • FIG. 19 is a block diagram of a network interface card for use with packet-merging network hubs
  • FIG. 20 is a block diagram of a further exemplary embodiment of a network hub in accordance with the present invention.
  • FIG. 21 is a block diagram of a plurality of packet-merging network hubs cascaded together according to the present invention.
  • Networking principles of the present invention are exemplified by the embodiments illustrated in the drawings. These principles focus on a two-channel approach to communicating data within a network environment. Exemplary embodiments of these networking principles may take different forms, many of which are illustrated in the drawings. However, the disclosure herein will detail specific preferred embodiments of this novel approach to networking and will not provide an exhaustive description of all embodiments within the scope of the networking principles set forth herein and in the appended claims.
  • the Network Referring to the drawings, particularly to FIG. 1, a network 10 implemented according to the principles of the present invention is shown.
  • the network 10 utilizes two separate transmission channels to communicate data within the network 10.
  • One of the transmission channels sequentially routes data issued from computers in the network 10 through the computers in the network.
  • the other transmission channel broadcasts this routed data to all the computers in the network 10.
  • data generated by computers is typically in the form of data packets.
  • Each data packet includes not only the data generated by the computer but may also include addressing information, error detection information, and so on, so that each packet may be an addressed data packet.
  • the network includes a plurality of computers 12, preferably being capable of generating data, in communication with each other via communication media.
  • Each of the computers 12 is provided with a network interface card (NIC) 14.
  • NIC network interface card
  • a network hub 16 interconnects the plurality of network interface cards 14, thereby interconnecting the computers 12 of the network 10.
  • the network hub 16 also communicates via a broadcast output 18 thereof with a broadcast input 20 of each of the network interface cards 14.
  • the interface between the computer 12 and the network interface card 14 allows the computer 12 to generate data packets and to provide these generated data packets (i.e., local data packets) to the network interface card 14 for transmission to the network 10.
  • the interface also allows the network interface card 14 to provide the computer 12 data packets received from the network or from a remote source (i.e., remote data packets).
  • the computers 12 are connected in the form of a chain of computers beginning with a first computer (i.e., Computer 1) and ending with a last computer (i.e., Computer N) in the chain.
  • a first computer i.e., Computer 1
  • Computer N a last computer
  • data generated by Computer 1 is sent to the network hub 16 which in turn sends this Computer 1 data to Computer 2.
  • Computer 2 may also generate and send its own local data to the network hub 16.
  • the network hub 16 then sends the Computer 1 and Computer 2 data to Computer 3.
  • This sequential linking of the computers 12 continues through to the last computer in the chain, i.e., Computer N.
  • the last computer is not linked back to the first computer.
  • Computer N receives remote data packets from the network hub 16, which remote data packets are data packets generated by any one of or all of the computers preceding Computer N in the network (i.e. , Computer 1 through Computer N— 1). Computer N sends these remote data packets back to the hub 16, and may add local Computer N data packets to the chain of data packets. Therefore, data packets received by the network hub 16 from the last computer in the chain of computers, i.e., Computer N, are the data packets generated by any and all the computers 12 in the network 10, which may be called "network data packets.” The network hub 16 then broadcasts these network data packets to all the network interface cards 14 in the network 10. Each of the network interface cards 14 then in turn provides these network data packets to the computer 12 connected thereto. Accordingly, all the computers 12 in the network 10 receive any data packet generated by any of the computers 12 in the network 10.
  • the networking principles of the present invention may be described as a network in which (1) data packets are sequentially gathered from the computers in the network on a reception or packet assembly medium and (2) the sequentially gathered data packets are delivered to all the computers in the network at the same time on a delivery or packet broadcast medium.
  • the network 10 includes a plurality of nodes or stations 22 (each station 22 preferably including a computer and a network interface card) through which data packets issued by the stations 22 are sequentially chained and routed on a reception medium 24.
  • the last station 22, i.e., Station N provides the sequentially chained and routed data packets to a broadcast unit 26 which broadcasts these data packets from the last station to all the stations 22 on a delivery medium 28, shown in dashed line.
  • the term "computer” is used to described any electronic device capable of being connected within a network, including computer workstations, file servers, computer peripherals (which may not be capable of generating data to send over the network), and so on.
  • the term "station” is used to describe a computer coupled with a network interface card.
  • the term "communication media” encompasses all media forms used to communicate data, including, for example, coaxial cable, twisted-pair cable, fiber optical systems, infrared systems.
  • the Network Hub The Network Hub
  • the network hub 16 may function as a media access unit by configuring the communication media of the network 10 into a reception or packet assembly channel for sequentially routing data packets through the last computer in the network and a delivery or packet broadcast channel for broadcasting data packets routed through the last computer in the network to all the computers in the network.
  • the network hub 16 includes a plurality of input-and-output pairs 30.
  • Each input-and-output pair 30 has an input 32 and an output 34 paired together.
  • Each input-and- output pair 30 communicates with the network interface card of a station or computer of the network 10 via the delivery channel 24 as described above.
  • the input 32 of each input-and- output pair 30 is connected to the output 34 of another input-and-output pair 30, shown by connection 36. Accordingly, data packets issued by a first station are received by the input 32 of a first input-and-output pair 30 and are then provided to the output 34 of a second input-and- output pair 30, which data packets are then sent to a second station.
  • the input-and-output pair 30 communicating with the last station in the network, i.e., Station N, provides data packets received by the input 32 thereof to a broadcast amplifier 38.
  • the broadcast amplifier 38 then provides these data packets to the broadcast output 18 to be broadcast to all the stations via the delivery channel 28.
  • the Network Interface Card A network interface card implemented according to the principles of the present invention is shown in FIG. 4.
  • the network interface card 14 includes a remote input 40 and a remote output 42, each of which is connectable to the reception medium 24.
  • the remote input 40 receives data packets from one of the input-and-output pairs 30 of the network hub 16 and provides these data packets to the remote output 42.
  • the remote output 42 in turn provides the data packets to the next input-and-output pair 30 sequentially in the hub 16.
  • the network interface card 14 further includes a local input 44 which is connectable to the computer 12.
  • the local input 44 is in communication with the remote output 42.
  • the local input 44 receives data packets from the computer to which it is connected and provides these data packets to the remote ou ⁇ ut 42 for transmission to the network hub 16 on the reception medium 24.
  • the remote input and output 40 and 42 and the local input 44 are concerned with the gathering, merging, and routing of data packets in the network on the first transmission medium 24.
  • the network interface card 14 further includes at least one but preferably two broadcast inputs 46 and a local output 48 in communication with the inputs 46.
  • the first broadcast input 46 is connectable to the delivery medium 28 of the network, receives the broadcast network data packets from the broadcast output 18 of the network hub 16, and provides the broadcast data packets to the local output 48.
  • the local output 48 then provides the network data packets to the computer to which it is connected.
  • the network interface card 14 may include a broadcast processor 49 connected between the broadcast inputs 46 and the local output 48.
  • the broadcast processor 49 receives data packets from the broadcast inputs 46 and determines which of the data packets are addressed to the local computer.
  • the broadcast processor 49 then transmits the data packets addressed to the local computer to the local output 48.
  • the broadcast processor 49 monitors data packet traffic and captures only those data packets fulfilling the address requirement programmed by the local computer.
  • the network interface card 14 may additionally include a memory device for storing address, as well as other, information.
  • the broadcast processor 49 may toggle the connection between the broadcast inputs 46 and the local output 48. This may take place under software control from the local computer. Further, the computer may decide which of the broadcast inputs 46 to monitor and receive data from under different circumstances.
  • each of the network interface cards 14 merges local data generated by the computer 12 with any remote data being received at the remote input 40.
  • the network interface card 14 in accordance with the present invention may also include a merging processor 50.
  • the merging processor 50 is in communication with the local input 44, the remote input 40, and the remote output 42.
  • the merging processor 50 alternatingly transmits to the remote output 42 local data packets received at the local input 44 and remote data packets received at the remote input 40.
  • the merging processor 50 upon receiving local data from the local input 44, creates the local data packet to be sent out to the network via the reception medium 24.
  • the preferred embodiment of the network interface card 14 also includes a local storage device 52 and a remote storage device 54.
  • the local storage device 52 is connected between the local input 44 and the merging processor 50.
  • the local storage device 52 receives data packets received at the local input 44 from the computer and provides these data packets to the merging processor 50.
  • the remote storage device 54 is connected between the remote input 40 and the merging processor 50.
  • the remote storage device 54 receives data packets received at the remote input 40 and provides these data packets to the merging processor 50.
  • the merging processor 50 detects the change in status of an "empty" flag in the local storage device 52.
  • the local data packets are then immediately released to the merging processor 50 (block C) and transmitted to the network (block D) via the remote output 42.
  • the merging processor 50 is currently busy transmitting remote data packets (path E)
  • the local data packets are then stored in the local storage device 52 (block F) until the merging processor 50 has transmitted the remote data packets to the network and is no longer "busy,” at which time the local data packets are release by the local storage device 52 (block C) and transmitted to the network (block D).
  • the release and transmission of local data packets is preferably temporarily suspended in order to transmit the remote packets through the network interface card 14 first, with any remaining local data packets transmitted thereafter. In other words, the transmission of remote data has priority over the transmission of local data.
  • the remote packets are loaded into the remote storage device 54 (block H) and any yet unreleased (and untransmitted) local data packets are stored in the local storage device 52 (block I) so that the remote packets may be released to the processor (block J) and transmitted to the network (block K).
  • any remaining unreleased local packets may then be released and transmitted (path L and blocks C and D).
  • This priority principle is alternatively illustrated in FIG. 5B.
  • local data packets are received at the local input 44 (block M) and loaded into the local storage device 52 (block N).
  • remote data packets are received at the remote input 40 (block O) and loaded into the remote storage device 54 (block P). If local data packets are received and the merging processor 50 is busy with remote data packets, the local packets are then stored in the local storage device 52 (block Q). If the processor 50 is not busy with remote packets, then the local packets are released to the processor 50 and transmitted to the network until remote data packets are received (path S).
  • the processor 50 When remote packets are received (block O), the processor 50 will transmit any released local packets to the network (block T), and any unreleased local packets are stored in the local storage device 52 (block U). Thereafter, the received remote data packets are released to the merging processor 50 and transmitted to the network (block W). After all the remote packets have been transmitted (path X), the release and transmittal of the local data packets may continue (block R). This entire process takes place at high speeds. If no local packets are being released and transmitted to the network and remote data packets are received at the remote input 40 (block O) and loaded into the remote storage device 54 (block P), then the merging processor 50 detects the change in status of an "empty" flag in the remote storage device 54. The remote data packets are then immediately released to the processor 50 (block V) and transmitted to the network (block W) via the remote output 42.
  • the networking principles of the present invention may be implemented in many various network topologies.
  • One such topology is a mirrored-server network which is a network having two file servers, one mirroring the other. Both file servers are configured to contain the same data, i.e., to "mirror" one another. Therefore, if one of the file servers crashes, the entire system will continue operating on the still- functioning file server.
  • Mirrored-server topologies are typically implemented in client/server networks.
  • a client/ server network has a plurality of user workstations (or “clients”) and a central file server (or “server”) to which the workstations are connected. This is also called server-based networking.
  • clients user workstations
  • server central file server
  • This is also called server-based networking.
  • workstations in a client/ server network typically only address data packets to the file servers, with the file servers addressing data packets to the workstations.
  • the network 56 includes a plurality of workstations 22 (each workstation 22 preferably including a computer and a network interface card) and a pair of file servers 58 connected together with communication media.
  • Each of the workstations 22 generates data packets intended for delivery to the file servers 58
  • each of the file servers 58 generates data packets intended for delivery to the workstations 22.
  • the communication media through which the workstations 22 and the file servers 58 communicate includes a plurality of channels, e.g., Channels A to E, as described below.
  • Each workstation 22 (except the last workstation in sequence, i.e., Workstation N) delivers data packets to the next workstation 22 in sequence on Channel A, and each workstation 22 (except the first workstation in sequence, i.e. , Workstation 1) receives data packets from the previous workstation 22 on Channel B.
  • Channels A and B are thus configured in a packet assembly channel.
  • Workstation N delivers these sequentially routed data packets to a broadcast unit (on Channel A).
  • the broadcast unit 60 broadcasts data from the workstations 22 to both of the file servers 58 on Channel C which is a workstation broadcast channel. Therefore, both file servers 58 receive the same data packets.
  • the file servers 58 each generate data packets, the same data packets in a mirrored- server configuration, to be delivered to the workstations on separate broadcast channels.
  • File Server 1 provides a first server broadcast unit 62 with addressed data packets which then broadcasts these File Server 1 data packets to the workstations 22 on Channel D which is a first server broadcast channel.
  • File Server 2 provides a second server broadcast unit 64 with addressed data packets which then broadcasts these File Server 2 data packets to the workstations 22 on Channel E which is a second server broadcast channel.
  • Each of the workstations 22 then receives the File Server 1 and 2 data packets on a first and a second server broadcast input connected to Channels D and E, respectively. Accordingly, if one of the file servers 58 crashes, the network 56 will continue to function on the still-functioning file server 58.
  • Each of the workstations 22 is then able to toggle between the first and second server broadcast inputs to determine which data packets to receive.
  • the workstations 22 may address data packets for delivery to another workstation.
  • each of the file servers 58 may merge data packets generated thereby with the workstation data packets received on Channel C, which is shown by the dashed line in each file server 58.
  • the network hub 66 implemented in accordance with the present invention is shown.
  • the network hub 66 configured for a mirrored-server environment includes a plurality of inputs 32 and outputs 34 formed into input-and-output pairs 30 which sequentially assemble data packets generated by Stations 1 to N on Channels A and B of the communication media.
  • the hub 66 also includes a workstation broadcast output 68 for transmitting data packets which have been sequentially routed through the last workstation 22 in the network (i.e., Station N) to the file servers 58.
  • the mirrored-server network hub 66 has a first server broadcast input 70 connected to the first file server 58 (i.e., File Server 1) and receives data packets generated by File Server 1 via Channel D.
  • the hub 66 also has a second server broadcast input 72 connected to the second file server 58 (i.e., File Server 2) and receives data packets generated by File Server 2.
  • the first and second server broadcast inputs 70 and 72 are respectively connected to the first and second server broadcast units 62 and 64.
  • the server broadcast units 62 and 64 Upon receiving data packets from the server broadcast inputs 70 and 72, the server broadcast units 62 and 64 respectively broadcast via Channels D and E these received data packets to the workstations 22.
  • the networking principles of the present invention are not limited to peer-to-peer networks or client/server networks but may be implemented in wide area networks as well.
  • a wide area network (WAN) 78 in accordance with the present invention is shown.
  • the WAN 78 includes a plurality of local area networks (LAN) 80 in communication with a remote source 82.
  • LAN local area networks
  • Each of the LANs 80 is comprised of a plurality of stations 22 in communication with a network hub 84 via communication media.
  • the remote source 82 which may be, for example, another network, a file server, a mail server, etc., communicates with the network hub 84 of each of the LANs 80.
  • the stations 22 of each LAN 80 sequentially gather data packets on Channels A and B of the communication media.
  • a LAN broadcast unit 86 of each hub 84 then broadcasts the data packets from the stations 22 to the stations 22 on Channel C. This is performed analogously to the exemplary embodiments described above.
  • the remote data packets are broadcast by the remote source 82 to each of the LANs 80 in the WAN 78.
  • a WAN broadcast unit 88 of each of the network hubs 84 receives the remote data packets from the remote source 82 and broadcasts the remote data packets to each of the stations 22 on Channel D of the communication media.
  • the network interface cards 14 and the network hub 16 do not require specialized communication media to implement a computer network according to the present invention.
  • Most types of cable available on the market are able to carry the reception and the delivery channels 24 and 28, which will be discussed further below.
  • a network 110 includes a file server 112 in communication with a plurality of workstations 114 via a network hub 16.
  • the file server 112 may include a processing unit or computer 118 which may inco ⁇ orate any number of devices including internal and external disk drives, microprocessors, and so on as is known in the art.
  • the workstations 114 each may include a processing unit or computer 120 which in turn may include internal and external disk drives, processors, and so on as is known in the art.
  • the network 110 further includes a plurality of network interface cards 14.
  • the data server 112 and each of the workstations 114 is provided with a network interface card 14.
  • the network interface card 14 may be housed inside the computers 118 and 120, as shown in FIG. 9, or provided as a separate, stand-alone unit in communication with the computers.
  • each of the network interface cards 14 may be in the form of a computer card with an interface bus 124 for connection with the respective computer 118 and 120.
  • the various connections between the network interface card 14, the network hub 16, and the processor units 118 and 120 may be made with known cables, which will be discussed in detail below.
  • the network interface card 14 preferably has three network inputs B, C, and D and one network output A, each in the form of cable pairs.
  • Input B is a remote input while inputs C and D are system broadcast inputs.
  • Output A is a combined output of the network interface card 14, preferably comprising cable pair 1 of the connector. Each of these connections has a specific and designated purpose in the wide bandwidth technology of the present invention.
  • the network interface card 14 may include a remote input first-in-first-out (FIFO) memory device 126 in communication with remote input B, and a local data FIFO memory device 128 in communication with the processor of the server 112 or the respective workstation 114 at which the network adapter 14 is located.
  • the network adapter 14 may further include a microprocessor or state machine 130 in communication with the FIFO memory devices 126 and 128.
  • the combined output A is output from the state machine 130. Accordingly, data packets that are input to the network interface card 14 through remote input B are combined in the state machine 130 with locally transmitted data from the local FIFO 128 and then sent to the hub 16 via the combined output A.
  • Ethernet-type schemes synchronize data transmissions from one computer to other computers on the local area network through a technique of data collision detection and recovery.
  • contention-type networks a considerable percentage of network bandwidth can be lost through data collisions when the network is under heavy utilization.
  • the networking technology of the present invention is a contentionless protocol. Data collisions are prevented by a loop-through approach which is accomplished in the network interface card 14 or hub 16. Data packets from other computers or workstations 114 enter the adapter 14 through remote input B. These incoming packets are stored in the remote input FIFO memory device 126. The remote input memory device 126 is sized with adequate depth to enable the temporary storage of the largest packet supported by the network 110. Local data, that is, data generated by the workstation 114 at which the specific network adapter 14 is located, to be transmitted is loaded into the local data FIFO memory device 128 through the interface bus 124 with the computer unit 118 or 120. The task of managing the synchronization of packets to be transmitted is accomplished by the microprocessor or state machine 130.
  • the state machine 130 detects the receipt of the packet of data by detecting a change in the status of the "empty" flag of the FIFO memory device 126, thereby signaling the beginning of the arrival of an incoming data packet. Upon such detection, the state machine 130 immediately begins transmitting, via combined output A, the incoming packet. In those instances when the local machine or workstation has created a packet of data for transmission and where the empty flag of the remote input FIFO memory device 126 indicates that no remote packet of data is being received, the state machine 130 then transmits a local packet at the local memory device 128 via the combined output A.
  • a remote packet of data is received by the state machine 130 from the remote FIFO 126 while the state machine 130 is in the process of transmitting a local packet of data from the local FIFO 128, the remote packet is stored in the remote input FIFO 126 and is transmitted immediately upon completion of the transmission of the local packet.
  • the combined output A preferably transmits at the same data rate as remote input B and as the remote input FIFO has the capacity to store an entire incoming packet of data, data overflow does not occur.
  • local packets of data are insertable into the data stream to be output from output A.
  • System broadcast inputs C and D have a special and unique function within the wide bandwidth network 110.
  • the adapter 14 has the ability of inputting data through either input C or D, one at a time, selectable under software control (stored on and activated by, for example, the processor units 118 or 120). As will be shown later, the two system broadcast inputs C and D input data to the local machine (i.e. , the server 112 or workstation 114 at which the adapter is located) and can be used in a diversity of methods depending on the requirements and restraints of the individual installation.
  • the local machine i.e. , the server 112 or workstation 114 at which the adapter is located
  • the wide bandwidth network interface card 14 is shown installed in a LAN configuration.
  • one of the workstations 114 for example, workstation 1, initiates a "chained" data stream.
  • Local data generated at workstation 1 is transmitted over the network 110 by workstation 1 and travels to the hub 16, where the workstation 1 local data is rerouted or "combined" to remote input B of the network adapter 14 located at workstation 2.
  • workstation 2 synchronizes the transmission of workstation 1 data with its own local data transmission, sending the combined output of workstation 1 and 2 local data to the hub 16 via combined output A where this combined data is chained to workstation 3.
  • Workstation 3 in turn synchronizes the transmission of workstation 1 and 2 data with its own local data transmission, sending the combined output of workstation 1, 2, and 3 local data to the hub 16 via combined output A.
  • transmissions from the combined output A of workstation 3 contain the combined queries or transmissions of all the workstations 114 of the network 110, which workstations 114 number three in this example.
  • the workstation 3 transmissions are in turn chained through the hub 16 to the data server 112.
  • the hub 16 includes a system broadcast amplifier 132 which receives the combined transmissions from the data server 112 and outputs the system broadcast transmission D.
  • An exemplary process for addressing data packets may include the following steps. Prior to the transmission of each packet, the computer loads into the network interface card 14 the destination and source MAC addresses. A MAC address is a unique number assigned to every network interface card 14. The first 48 bits of the MAC address are the vendor's ID assigned to that vendor by the I.E.E.E.
  • the last 48 bits are a distinguishing number for each board assigned by the manufacturer of the board.
  • the sender's source address is stored in permanent memory on the network interface card 14.
  • the computer reads the MAC address of the network interface card 14 connected thereto. The computer then loads the address into a register at an appropriate time. From the data loaded in the register, the network interface card 14 creates and sends the addressed data packet.
  • An exemplary process for receiving data packets may include the following steps.
  • the network interface card 14 receives an address from the computer for the desired packet. Most of the time, this address would be the local MAC address of the particular network interface card 14. At other time, other addresses may be used. For example, a promiscuous mode utilized by network-monitoring products is programmed to receive all packets with all labels. There may be another address for broadcast packets. After the network interface card 14 is programmed with a receive address, the network interface card 14 then automatically monitors packet traffic and captures only those packets which fulfill the address requirement of the computer. Returning to the embodiment shown in FIG.
  • system broadcast or WideCastTM is used to refer to a signal or transmission which is simultaneously sent to a number of computers, workstations, servers, etc.
  • system broadcast output D is delivered simultaneously as system broadcast input D to each of the workstations 114 and to the server 112.
  • the system broadcast output D signal is delivered to every computer connected to the hub 16.
  • System broadcast output D is preferably transmitted over pair four of the unshielded twisted pair (UTP)-5 cable connecting each computer unit 118 and 120 (i.e., the adapter 14) to the hub 16.
  • UTP unshielded twisted pair
  • the network interface card 14 which monitors all of the packet transmissions on the channel, selects and retrieves those packets of data addressed to the local computer at which the particular network interface card 14 is located. It is by this method that round-trip communications are accomplished in the wide bandwidth environment of the present invention.
  • FIG. 11 there is shown a wide bandwidth network 134 in which a pair of hubs, a server hub 136 and a workstation hub 138, are provided, as well as more than one server 112.
  • the transmissions of the workstations 114 have been separated from the transmissions of the servers 112.
  • the bandwidth of a single hub 16 is limited to the combined transmission of all workstations 114 and servers 112.
  • FIG. 12 illustrates a simplified wiring diagram of a wide bandwidth hub 144 according to the present invention.
  • the wide bandwidth hub 144 provides a simple method of connecting up to 24 computers into a LAN by tying together sequentially inputs and outputs of the hub 144 (respectively represented by reference numerals 1 to 24), thereby connecting the combined output of one of the computers of the LAN to the remote input of another one of the computers of the LAN.
  • FIG. 13 illustrates another exemplary embodiment of the wide bandwidth network technology of the present invention.
  • a wide area network 150 includes a plurality of LANs 152 each including a combination of workstations and servers.
  • a hub 154 and a corresponding number of routers 156 are connected into a backbone configuration.
  • Each of the routers 156 has a network interface card 14 analogous to that described above in reference to FIG. 10.
  • the hub 154 includes a system broadcast amplifier 158 which transmits the combined data stream of all of the LANs 152 to each of the routers 156 connected to the hub 154.
  • FIG. 14 Another implementation of the wide bandwidth technology of the present invention is illustrated.
  • the network of Fig. 14 is similar to that of FIG. 13 with the addition of access to the Internet, represented by reference numeral 160, via an Internet server 162 including a network adapter 14.
  • a high-speed Internet interface such as a Tl or a T3 is coupled, full speed, all the way to the desktop.
  • FIG. 15 illustrates a wide area network 164 in accordance with the present invention.
  • the wide bandwidth components are connected into wide area networks in which system broadcast D is utilized for distribution of local area network traffic and in which system broadcast C provides a high-speed, enterprise-wide channel which greatly simplifies E-mail installations and accelerates wide area communications via a mail server system 166.
  • each of the networks includes a combination of servers and workstations, as well as a hub 168 including a system broadcast D amplifier 170 corresponding to combined data streams of the local area network and a system broadcast C amplifier 172 corresponding to the mail server 166.
  • a hub 168 including a system broadcast D amplifier 170 corresponding to combined data streams of the local area network and a system broadcast C amplifier 172 corresponding to the mail server 166.
  • the network interface card 14 monitors system broadcast input C to retrieve locally addressed mail and message packets transmitted from the mail server system 166.
  • the implementation of the wide bandwidth network 174 of the present invention illustrated in FIG. 16 depicts two servers 112 operating in parallel or in a "mirrored" mode. All of the queries to the servers 112 are delivered simultaneously to both servers 112 over system broadcast ou ⁇ ut C from a server hub 176.
  • the ou ⁇ ut of file server 112 is delivered to each of the workstations 114 over system broadcast channel C from a workstation hub 178, whereas the ou ⁇ ut transmissions of file server 112 are delivered to each of the workstations 114 via the system broadcast channel D of the workstation hub 178.
  • both hubs 176 and 178 operate in parallel, processing requests and responding to the workstations 114.
  • both servers 112 are completely independent and redundant, even down to the cable pair over which the server data is delivered to the workstations 114.
  • the workstation 114 could then independently switch from input channel D to input channel C and continue processing with the still functioning server 112.
  • This wide bandwidth network 180 includes a video source 182 connected to a network hub 184 via input port B.
  • a second video source 186 is connected to the next port, continuing down the chain as discussed in reference to FIG. 17.
  • packets of data can automatically be sent over the network 180 in synchronization with the demands of the video capture device as shown.
  • the synchronization of data packets at the workstations 114 can be accomplished utilizing FPGA technology such as the Cypress 384, 385, and 387 FPGAs. These devices are fast and user programmable. They can also simultaneously perform data encryption functions as disclosed in U.S. patent application serial No. 08/430,942 filed on April 26, 1995, which application is inco ⁇ orated herein by reference and by appending a copy of the application hereto.
  • data is transmitted serially.
  • One way to accomplish this serial transmission is by transferring the data into the parallel input of a device such as the Cypress CY7B923 which performs 8B/10B coding and which creates the differential high-speed serial output. Details on various coupling and termination techniques for serial communication are discussed in the CY7B923 and 933 Data Book.
  • Wideband data is transmitted from the server 112 to the workstations 114 at a bit rate of preferably 333 Mbps over each twisted pair of UTP-5 cables.
  • Eight-bit data is converted into a 10-bit format to maintain clock synchronization, to provide a method of hardware error detection, and to enable the transmission of control characters.
  • the decoding of 10-bit data back to its original 8-bit format on the receiving side results in a useful data rate throughput of 267 Mbps per cable pair or 33 megabytes per second (MBps). Since three or the four twisted pairs of the UTP-5 cable are utilized to transfer data by the wide bandwidth technology, the total data rate is preferably 1.0 Gbps.
  • the fourth pair is utilized for data synchronization.
  • the preferred technical specifications of the wide bandwidth technology of the invention are outlined in TABLE 1.
  • Bit Error Rates IO "12 or better Wideband transmission distances over UTP-5 cable are 100m (330 ft). Transmission distances of other types of cabling and fiber are presented in TABLE 2.
  • the trappings on the cables range from metal Ethernet T-connectors to the simple plastic modular plugs used with unshielded twisted pair (UTP) wire.
  • Some Ethernet cards have connectors for coaxial cable, and others provide a 15-pin socket for more complex external transceivers for fiber optic and other types of cables.
  • Token Ring cards have a 9-pin connector for shielded twisted pair (STP) wire.
  • STP shielded twisted pair
  • UTP wire is becoming increasingly popular for both Ethernet and Token Ring cards.
  • RJ-45 simple plastic rectangular jack
  • Table 2 also provides data transmission distances for systems compensated with passive equalization.
  • Wideband networking is a sophisticated yet simple approach to increasing data transmission bandwidth in local and wide area networks over existing cabling.
  • Wideband utilizes the basic technology of ATM but with modifications which make it more readily compatible with existing application software and the OSI seven-layer model. Through all its various configurations, it provides a versatile alternative in high-speed networking.
  • the network interface cards do not need to use copper cables.
  • the network interface cards may be configured to read pulses of laser light sent over fiber optic cables, pulses of infrared light sent through the air, or signals imposed on radio waves.
  • Many installations may use copper cables for most connections and intermix fiber optic or wireless alternatives to reach special stations.
  • a network 200 includes a packet- merging hub 202 in communication with each of the computers 12 via a respective network interface card 204.
  • the data packets are sent from the computers 12 to the packet- merging hub 202 to be merged into a stream of data packets which is then broadcast to all of the computers 12 in the network 200.
  • the packet- merging hub 202 includes a plurality of ports 206 each for connecting with and receiving data packets from a respective one of the computers via a dedicated communication medium, such as a data channel 208.
  • Each of the ports 206 is paired with and connected to a storage device, such as a first-in, first-out (FIFO) device 210.
  • a storage device such as a first-in, first-out (FIFO) device 210.
  • Each of the FIFO devices 210 receives and temporarily stores data packets received from one of the computers 12 via the port 206 paired therewith.
  • the packet-merging hub 202 further includes a merging processor 212, such as a Cypress 382 FPGA, connected to each of the FIFO devices 210.
  • the merging processor 212 sequentially merges data packets received at each of the ports 206 into a stream of data packets.
  • the steam of data packets includes all of the data packets generated by each of the computers 12.
  • the packet-merging hub 202 includes an internal bus 214 on which the stream of data packets are delivered to a broadcast ou ⁇ ut 216 connected to a broadcast channel 218.
  • the internal bus 214 is substantially a packet- merging channel provided in the network hub 202.
  • the packet-merging hub 202 then broadcasts the stream of data packets to each of the computers 12 via the broadcast output 216 and the broadcast channel 218.
  • the packet-merging hub 202 preferably includes an amplifier 220 for amplifying the stream of data packets prior to broadcast on the broadcast channel 218.
  • each of the network interface cards 204 includes a local data input 222 connected to one of the computers 12 and a local data ou ⁇ ut 224 connected to one of the data channels 208.
  • the data input 222 receives data packets generated by the computer 12 connected thereto, and the data output 224 provides the generated data packets to the network hub 202 via the data channel 208.
  • Each of the interfaces 204 also includes at least one broadcast input 226 connected to a broadcast ou ⁇ ut 228 with a broadcast processor 230 provided therebetween.
  • the broadcast input 226 is connected to the broadcast channel 218 and receives the stream of data packets.
  • the broadcast processor 230 determines which of the data packets are addressed to the local computer 12 and provides such data packets to the broadcast output 228 for delivery to the local computer 12 for further processing.
  • Each of the interfaces 204 preferably includes a local storage device 231 connected between the local data input 222 and the local data output 224.
  • the local storage device 231 is able to temporarily store data packets generated by the local computer 12 when the FIFO device 210 corresponding thereto is filled or nearly filled to capacity during peak operating conditions of the network hub 202.
  • data packets temporarily stored in the FIFO devices 210 may have to queue until the merging processor 212 is able to merge the temporarily stored data packets into the stream of data packets.
  • the FIFO devices 210 have a limited storage capacity, if data packets are being stored in one of the FIFO devices 210, waiting to be merged, and if the computer associated therewith continues to generate and transmit data packets to the queueing FIFO device 210, then there is the possibility that the FIFO device 210 will reach its storage capacity before the merging processor 212 is able to relieve the FIFO device 210 of the temporarily stored data packets.
  • the FIFO device 210 it is preferable for the FIFO device 210 to transmit a status signal or flag to the interface 204, indicating that the FIFO device 210 is nearing or is at its storage capacity, or that the FIFO device 210 has su ⁇ assed a predetermined percentage of its total storage capacity, e.g., 95%.
  • the interface 204 Upon receipt of the status signal, the interface 204 temporarily stores already-generated data packets in the local storage device 231 until the FIFO device 210 has unloaded the stored data packets into the stream of data packets. It is preferable to send the status signal from the network hub 202 to the interface 204 on the same data channel 208 connecting the port 206 with the data output 224.
  • the merging processor 212 preferably monitors each of the FIFO devices 210 to determine which of the FIFO devices 210 are nearing storage capacity and, responsively, to merge the data packets from such FIFO devices 210 into the stream of data packets expeditiously.
  • the data output 224 of the interface 204 is included in a data port connected to the data channel, which data port also includes a signal input for receiving the signal from the network hub 202.
  • each of the data channels 208 is able to deliver data packets from the interfaces 204 to the network hub 202 and is able to deliver the capacity-indication signal from the network hub 202 to the interface 204, which two-way communication is indicated by the double-headed arrows of the data channels 208 in FIG. 20.
  • the data channels 208 are preferably cable pairs, with one cable dedicated to the transmission of data packets and the other cable dedicated to the transmission of the status signal.
  • each of the interfaces 204 may also include a second broadcast input 232 connected to another broadcast channel 234 and the broadcast processor 230.
  • the broadcast channel 232 may be connected to a remote source, such as the Internet, a video server, or an electronic mail system, for receiving a remote stream of data packets.
  • the broadcast processor 230 receives the remote stream of data packets, determines which of the remote data packets are addressed to the local computer, and then provides such data packets to the broadcast output 228 for deliver to the local computer 12.
  • each of the network hubs 202 may be serially connected into a multi-hub network 240.
  • each of the network hubs 202 further includes an expansion port 242 which is connected to the internal bus 214 of the hub 202.
  • the expansion port 242 is provided with the same stream of data packets as that provided to the broadcast ou ⁇ ut 216 of each of the hubs 202.
  • the expansion port 242 of a first network hub is connected to one of the ports 206 of a second network hub, thereby providing the stream of data packets from the first network hub to the second network hub, indicated by reference numeral 244 in FIG. 21.
  • the merging processor 212 of the second network hub 202 then merges the stream of data packets 244 from the first network hub 202 with the data packets received at its other ports 206, generating a combined stream of data packets.
  • the expansion port 242 of the second network hub may then be connected to one of the ports 206 of a third network hub 202, thereby providing the combined stream of data packets to the third network hub 202, as indicated by reference numeral 246.
  • the merging processor 100 of the third network hub 202 then merges the combined stream of data packets 246 with data packets received at its other ports 206.
  • the stream of data packets merged by the third network hub 202 may then be broadcast to all of the computers in the network 240 via the broadcast channel 218.
  • the network 240 with cascaded network hubs 202 may also be connected to a remote source 248, such as the Internet, a wide area network, or an electronic mail system, for receiving a remote stream of data packets 250 from the remote source 248.
  • a remote source 248 such as the Internet, a wide area network, or an electronic mail system
  • each of the network hubs 202 preferably include a first broadcast input 252 and a second broadcast input 254.
  • the remote source 248 is connected to the first broadcast input 252.
  • the second broadcast input 254 may then be connected to another remote source for further expansion of the network 240.
  • the expansion port 242 of the third network hub 202 receiving the combined stream of data packets 246 is connected to the second broadcast input 254 by a jumper cable 256.
  • the third network hub 202 is able to broadcast to all of the computers in the network 240 both the stream of data packets from all the computers in the network 240 and the remote stream of data packets 250 from any remote source 248 that may be connected to the network 240.
  • Each of the network hubs 202 may also include first and second broadcast amplifiers 258 and 260 for amplifying the respective streams of data packets to be broadcast over the two broadcast channels 218 and 234.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Data Exchanges In Wide-Area Networks (AREA)

Abstract

Un réseau informatique (200) comporte un concentrateur de réseau (202) qui connecte plusieurs ordinateurs (12) entre eux. Des paquets de données générés par les ordinateurs (12) sont fusionnés de manière séquentielle de sorte qu'une suite de paquets de données soit formée, suite qui est ensuite envoyée à tous les ordinateurs (12) du réseau (200). Le concentrateur (202) comporte plusieurs ports (206) connectés chacun à un des ordinateurs (12) respectifs, et recevant les paquets de données générés par l'ordinateur (12). Le concentrateur (202) comporte également plusieurs mémoires (210) connectées respectivement aux ports (206). Chaque mémoire (210) reçoit les paquets de données reçus par le port (206) et les mémorise temporairement. Un processeur de fusion (212) est connecté à chaque mémoire (210) et fusionne séquentiellement les paquets de données mémorisés temporairement par les mémoires (210), sous forme de suite de paquets de données. La suite de paquets de données est ensuite envoyée à chaque ordinateur (12) du réseau (200). Le concentrateur (202) peut communiquer avec chaque ordinateur (12) pour indiquer l'état de capacité des mémoires (210).
EP96940417A 1995-11-13 1996-11-12 Technologie pour reseau a grande largeur de bande Withdrawn EP0861469A1 (fr)

Applications Claiming Priority (13)

Application Number Priority Date Filing Date Title
US55651895A 1995-11-13 1995-11-13
US556518 1995-11-13
US690735 1996-07-31
US08/687,777 US6115747A (en) 1995-11-13 1996-07-31 Computer network interface that merges remote data received from other computers with local data before transmitting the merged data to a network
US690676 1996-07-31
US08/690,676 US5793981A (en) 1995-11-13 1996-07-31 System for communicating data in a network using both a daisy chain link and separate broadcast links
US687777 1996-07-31
US688799 1996-07-31
US08/688,799 US6061730A (en) 1995-11-13 1996-07-31 Methods and apparatus for communicating data in computer networks with separate packet assembly and packet broadcast channels
US08/690,673 US5761433A (en) 1995-11-13 1996-07-31 System for communicating data in a network using both a daisy chain link and separate broadcast links
US08/690,735 US6076117A (en) 1995-11-13 1996-07-31 Packet merging hub system for sequentially merging received data in a network hub into data packets before broadcasting to a plurality of destination computers
US690673 1996-07-31
PCT/US1996/018196 WO1997020271A1 (fr) 1995-11-13 1996-11-12 Technologie pour reseau a grande largeur de bande

Publications (1)

Publication Number Publication Date
EP0861469A1 true EP0861469A1 (fr) 1998-09-02

Family

ID=27560130

Family Applications (1)

Application Number Title Priority Date Filing Date
EP96940417A Withdrawn EP0861469A1 (fr) 1995-11-13 1996-11-12 Technologie pour reseau a grande largeur de bande

Country Status (4)

Country Link
EP (1) EP0861469A1 (fr)
AU (1) AU7730396A (fr)
CA (1) CA2237406A1 (fr)
WO (1) WO1997020271A1 (fr)

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0447228A3 (en) * 1990-03-16 1993-01-07 Hewlett-Packard Company Data stream concentrator providing attribute data storage and graphics pipeline access
US5406559A (en) * 1992-11-02 1995-04-11 National Semiconductor Corporation Isochronous link protocol
US5361262A (en) * 1993-04-16 1994-11-01 Bell Communications Research, Inc. Estimated-queue, expanded-bus communication network

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See references of WO9720271A1 *

Also Published As

Publication number Publication date
CA2237406A1 (fr) 1997-06-05
AU7730396A (en) 1997-06-19
WO1997020271A1 (fr) 1997-06-05

Similar Documents

Publication Publication Date Title
US6115747A (en) Computer network interface that merges remote data received from other computers with local data before transmitting the merged data to a network
US6076117A (en) Packet merging hub system for sequentially merging received data in a network hub into data packets before broadcasting to a plurality of destination computers
US5761433A (en) System for communicating data in a network using both a daisy chain link and separate broadcast links
US5953340A (en) Adaptive networking system
US6510161B2 (en) Low latency shared memory switch architecture
US5574722A (en) Protocol independent switch
US6662254B1 (en) System architecture
US6169729B1 (en) 200 Mbps PHY/MAC apparatus and method
JP2537017B2 (ja) 通信ネットワ―ク相互接続装置
US5828475A (en) Bypass switching and messaging mechanism for providing intermix data transfer for a fiber optic switch using a bypass bus and buffer
US5764895A (en) Method and apparatus for directing data packets in a local area network device having a plurality of ports interconnected by a high-speed communication bus
JP4972206B2 (ja) パケット切替装置、および、このようなパケット切替装置を備えるローカル通信ネットワーク
US6272552B1 (en) Dual channel 100Base-T link for computer networks
JPH11145990A (ja) イーサネット（登録商標）互換ネットワークのトランク化
JPH1065740A (ja) ネットワーク・スイッチ及び該スイッチを利用したデータ・パケット伝送方法
US6925078B1 (en) Network adaptor driver with destination based ordering
US5349583A (en) Multi-channel token ring
EP2497031B1 (fr) Commutateur de réseau
EP2117174A1 (fr) Dispositif avec fonction de commutation Ethernet et connecteur Ethernet unique
US6061730A (en) Methods and apparatus for communicating data in computer networks with separate packet assembly and packet broadcast channels
US5793981A (en) System for communicating data in a network using both a daisy chain link and separate broadcast links
US6484213B1 (en) Adapting networking device for enhancing performance of a hybrid networking system
US6118782A (en) Asynchronous transfer mode network
WO1997020271A1 (fr) Technologie pour reseau a grande largeur de bande
Ocheltree et al. A comparison of fibre channel and 802 MAC services

Legal Events

Date Code Title Description
PUAI Public reference made under article 153(3) epc to a published international application that has entered the european phase

Free format text: ORIGINAL CODE: 0009012

17P Request for examination filed

Effective date: 19980615

AK Designated contracting states

Kind code of ref document: A1

Designated state(s): AT BE CH DE DK ES FI FR GB GR IE IT LI LU MC NL PT SE

STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: THE APPLICATION IS DEEMED TO BE WITHDRAWN

18D Application deemed to be withdrawn

Effective date: 20030603