Abstract: A method for synchronizing local clocks in a distributed computer network, wherein end systems and switches of the network executes the method as a synchronization state machine, which uses three different frame types. The states in the state machine belong to an unsynchronized or to a synchronized set of states. All end systems being configured as Synchronization Master periodically send coldstart frames in one of the unsynchronized states, all end systems being configured as Synchronization Master react to the reception of a coldstart frame by sending a coldstart acknowledgment frame a first timeout after the reception of the coldstart frame on all replicated communication channels. First timeout is reset when a consecutive coldstart frame is received before the coldstart acknowledge is sent, and all Synchronization Masters react to the reception of a coldstart acknowledgment frame by starting a second timeout and enter a synchronized state when the second timeout expires.

Abstract: The goal of the present invention is to improve the useful data efficiency and reliability in the use of commercially available ETHERNET controllers, in a distributed real time computer system, by a number of node computers communicating via one or more communication channels by means of TT ETHERNET messages. To achieve this goal, a distinction is made between the node computer send time (KNSZPKT) and the network send time (NWSZPKT) of a message. The KNSZPKT must wait for the NWSZPKT, so that under all circumstances, the start of the message has arrived in the TT star coupler at the NWSZPKT, interpreted by the clock in the TT star coupler. The TT star coupler is modified, so that a message arriving from a node computer is delayed in an intelligent port of the TT star coupler until the NWSZPKT can send it precisely at the NWSZPKT into the TT network.

Abstract: The aim of the present invention is that of establishing a fault-tolerant global time in a fault-tolerant communication system of a distributed real-time system. For this purpose, a fault-tolerant message switching unit is provided, which is composed of four independent switching units. These four independent switching units jointly establish a fault-tolerant time. The terminal systems are connected to a fault-tolerant message switching unit via two independent fail-silent communication channels, so that the clock synchronization and network connections are preserved, even if a part of the fault-tolerant switching unit or of a communication channel fails.

Abstract: The goal of the present invention is to improve the useful data efficiency and reliability in the use of commercially available ETHERNET controllers, in a distributed real time computer system, by a number of node computers communicating via one or more communication channels by means of TT ETHERNET messages. To achieve this goal, a distinction is made between the node computer send time (KNSZPKT) and the network send time (NWSZPKT) of a message. The KNSZPKT must wait for the NWSZPKT, so that under all circumstances, the start of the message has arrived in the TT star coupler at the NWSZPKT, interpreted by the clock in the TT star coupler. The TT star coupler is modified, so that a message arriving from a node computer is delayed in an intelligent port of the TT star coupler until the NWSZPKT can send it precisely at the NWSZPKT into the TT network.

Abstract: A system comprises a plurality of nodes, at least one of the plurality of nodes configured to insert, on a per-virtual link basis, a delay value into a dynamic delay field of a frame corresponding to the respective virtual link, wherein the dynamic delay value represents latency of frames of the respective virtual link. The system also comprises a switch having a plurality of ports, each port coupled to one of the plurality of nodes. The switch is configured to route frames received from the plurality of nodes to one or more of the plurality of nodes. At least one of the plurality of nodes is configured to store frames received from the switch in a buffer and to update the value in the dynamic delay field to reflect the end-to-end system delay.

Abstract: A communication system comprises a plurality of nodes and a switch having a plurality of ports, each port coupled to one of the plurality of nodes. At least one of the plurality of nodes is configured to transmit a first unregulated flow of frames associated with a first virtual link to a first port of the plurality of ports of the switch. The switch is configured to regulate the first unregulated flow of frames by buffering frames of the first unregulated flow in a first input queue associated with the first virtual link, applying traffic shaping parameters associated with the first virtual link to the frames in the first input queue associated with the first virtual link, and outputting the first regulated flow of frames to one or more output queues associated with one or more output ports based on the first virtual link.

Abstract: A communication switch comprises a plurality of ports and a processing unit configured to identify source ports and destination ports from the plurality of ports on a per virtual link basis. Each virtual link comprises one or more source ports via which frames are received from one or more source nodes and one or more destination ports via which received frames are forwarded to one or more destination nodes. For at least one virtual link, the processing unit is configured to accept more than one of the plurality of ports as source ports.

Abstract: A method for start-up of a network, including a number of nodes, which are connected via channels. The nodes exchange information in the form of messages via the channels. The transition phase of a synchronizing node from its initial phase to a synchronized phase is separated in a first integration phase and a second subsequent cold-start phase. A synchronizing node in the integration phase listens to messages being sent from nodes in the synchronized phase and only reacts to an integration message (i-frame) if the integration message is a valid message. Furthermore, a synchronizing node, wherein integration of the synchronizing node to a set of already synchronized nodes was not successful after a specifiable period, changes into the cold-start phase, in which a cold-start procedure of the node is extracted, wherein in the cold-start phase the node does not react to integration messages of a node in the synchronized phase.

Abstract: A method for start-up of a network, including a number of nodes, which are connected via channels. The nodes exchange information in the form of messages via the channels. The transition phase of a synchronizing node from its initial phase to a synchronized phase is separated in a first integration phase and a second subsequent cold-start phase. A synchronizing node in the integration phase listens to messages being sent from nodes in the synchronized phase and only reacts to an integration message (i-frame) if the integration message is a valid message. Furthermore, a synchronizing node, wherein integration of the synchronizing node to a set of already synchronized nodes was not successful after a specifiable period, changes into the cold-start phase, in which a cold-start procedure of the node is extracted, wherein in the cold-start phase the node does not react to integration messages of a node in the synchronized phase.

Abstract: The invention relates to a method for synchronizing local clocks in a distributed computer network, where said computer network consists of a number of end systems and at least two switches. Each end system is connected to at least two switches via bi-directional communication links. A configured subset of end systems and switches executes the method in form of a synchronization state machine. The state machine uses at least three different frame types. The states in the state machine are either said to belong to an unsynchronized set of states or belong to a synchronized set of states.

Abstract: The invention relates to a method for synchronizing local clocks in a distributed computer network, where said computer network consists of a number of components that are connected to each other via bi-directional communication links, characterized in that an a priori configured set of components of the network generates synchronization messages and sends these synchronization messages to the network. An a priori configured set of components consumes the generated synchronization messages. One or more message-relaying components in between a component that generates a synchronization message and the component that consumes the synchronization message will add a delay value to the synchronization message reflecting the delay imposed by the message-relaying component.

Abstract: The goal of the present invention is to improve the useful data efficiency and reliability in the use of commercially available ETHERNET controllers, in a distributed real time computer system, by a number of node computers communicating via one or more communication channels by means of TT ETHERNET messages. To achieve this goal, a distinction is made between the node computer send time (KNSZPKT) and the network send time (NWSZPKT) of a message. The KNSZPKT must wait for the NWSZPKT, so that under all circumstances, the start of the message has arrived in the TT star coupler at the NWSZPKT, interpreted by the clock in the TT star coupler. The TT star coupler is modified, so that a message arriving from a node computer is delayed in an intelligent port of the TT star coupler until the NWSZPKT can send it precisely at the NWSZPKT into the TT network.

Abstract: A method for start-up of a network, including a number of nodes, which are connected via channels. The nodes exchange information in the form of messages via the channels. The transition phase of a synchronizing node from its initial phase to a synchronized phase is separated in a first integration phase and a second subsequent cold-start phase. A synchronizing node in the integration phase listens to messages being sent from nodes in the synchronized phase and only reacts to an integration message (i-frame) if the integration message is a valid message. Furthermore, a synchronizing node, wherein integration of the synchronizing node to a set of already synchronized nodes was not successful after a specifiable period, changes into the cold-start phase, in which a cold-start procedure of the node is extracted, wherein in the cold-start phase the node does not react to integration messages of a node in the synchronized phase.