Method of broadcasting data by means of a data train

• US 5,483,520 A
• Filed: 10/20/1994
• Issued: 01/09/1996
• Est. Priority Date: 10/22/1993
• Status: Expired due to Term

• Current Assignee: Cegelec (VINCI SA)
• Original Assignee: Cegelec (VINCI SA)
• Inventors: Eychenne, Yves, Simatic, Michel
• Primary Examiner(s): Olms, Douglas W.
• Assistant Examiner(s): BLUM, RUSSELL
• Application Number:
  US08/326,202
• Time in Patent Office: 446 Days
• Field of Search: 370/16.1, 370/85.5, 370/85.12, 370/85.15, 371/11.1, 371/11.2
• US Class Current: 370/222
• CPC Class Codes: H04L 12/437 Ring fault isolation or reconfiguration for SDH/SONET ring networks H04J3/085

Abstract

A method of transmitting data between a set of data processing nodes interconnected by a communications network. Each node broadcasts data to the other nodes by causing a train of such data to travel over the network around a looped circuit or "ring" passing through each of said nodes, each node in the ring having a predecessor and a successor. A counter is provided in the data train, and on receiving a data train sent by its predecessor, each node recovers the counter from the received data train, increments the counter, and dispatches to its successor a new data train including the counter as incremented. It also stores the new data train with the incremented counter in its memory so that it is possible to implement a simple fault recovery procedure in the event of one of the nodes becoming faulty.

Claims

1. A method of transmitting data for a set of data processing nodes that are interconnected by a communications network and for the purpose of broadcasting data from each node to the other nodes by causing a train of such data to travel over the network around a looped circuit or ring that passes through said data processing nodes in succession in such a manner that each node in turn recovers, removes, and writes data in the train, each node having a predecessor and a successor on the ring and being capable of detecting a faulty state of its successor or of its predecessor;

wherein prior to writing new data to be broadcast in the train, each node begins by performing the steps of;

storing in an internal memory of the node the data of the train that the node has just received, including the value of a train header counter,

incrementing the value of this counter in the node'"'"'s internal memory, and

dispatching the train back over the network containing new data to be broadcast and with an incremented value in the header counter; and

wherein when a node detects that its predecessor on the ring is in a faulty state, the node performs the steps of;

undertaking a procedure of searching for a new predecessor on the ring,

establishing a new ring for the train,

receiving a train from its new predecessor, and

before proceeding to store in the node'"'"'s internal memory the data of the train that it has just received from its new predecessor;

comparing the value of the counter in the header of the received train with the value of the same counter as stored in the node'"'"'s internal memory,

dispatching over the network the train containing said new data if the value of the counter in the header of the received train is greater than the value of the counter as stored in the node'"'"'s internal memory, and

waiting to receive another data train if the value of the counter in the header of the received train is not greater than the value of the counter as stored in the node'"'"'s internal memory.

2. The method according to claim 1, in which whenever a particular node has detected that its predecessor is in a faulty state and has established a new ring for the train to travel, the predecessor of the faulty node dispatches over the network a train having the data and the counter value as stored in its internal memory.