TL;DR: In this paper, the authors describe a ring of point-to-point links for transport of fixed rate synchronous, fixed rate asynchronous data and variable rate data in a flexible format, where the allocation of capacity among variable width channels is revised block by block, and a transition period is defined to allow for ring latency.
Abstract: Examples of local communication systems are disclosed, based on a ring of point-to-point links, providing for transport of fixed rate synchronous, fixed rate asynchronous data and variable rate data in a flexible format. Different segments of the ring network can carry data at different bit rates, while remaining synchronised to a common frame rate and having a common control channel structure, for compatibility with earlier systems. Parallel channels are provided, either permanently or when required, for signalling errors of source data, data validity/padding, flow control. Parallel variable width channels are defined with free content (stream or packet). Null data symbols are defined for padding on a byte-by-byte basis. The allocation of capacity among variable width channels is revised block by block, and a transition period is defined to allow for ring latency. Calculations for allocations of capacity are performed during one block for the next block, locally at each source station, according to predetermined rules. Information as to bandwidth requirements is exchanged prior to the calculation via a special connection signalling channel and message format.
TL;DR: An analysis of self-timed pipelines and rings is presented, and the author defines terminology, qualitatively describes data and bubble flow, and defines the variables locally characterizing each stage in a ring in terms of those variables locally describing the stages.
Abstract: An analysis of self-timed pipelines and rings is presented. The author defines terminology, qualitatively describes data and bubble flow, and defines the variables locally characterizing each stage in a ring. Values for these variables can be found using the dependency graph analysis method presented. The author then determines the overall performance of self-timed rings in terms of those variables locally describing the stages. The performance is expressed on graphs showing regions defined by the number of stages and tokens. Separate equations give the ring latency and throughput within each region. Cross-sections graphing latency and throughput versus the number of stages and tokens are presented, and the implications of this analysis for self-timed ring design are discussed. >
TL;DR: In this article, a ring network is considered, where each node has its own clock to provide timing needed by equipment at that node, and each node performs its own measurement of propagation delay, comparing its measurement with that of the master node.
Abstract: An asynchronous, fiber optic, ring network includes a number of nodes where data enters and exits the network. Each of these nodes has its own clock to provide timing needed by equipment at that node. So that synchronous data can be transmitted between predetermined nodes, a master node provides timing information which may be used at any node to synchronize its clock. Timing information comprises the propagation delay around the ring (ring latency) as measured by the master node. This measure of propagation delay is transmitted by the master node as an information packet available at all nodes. Synchronization of any other node with the master requires that the other node perform its own measurement of propagation delay, compare its measurement with that of the master node, and make adjustments to its own clock that tend to decrease the difference between these measurements.
TL;DR: In this paper, a plurality of stations are used to strip expendable messages from a serial ring communication network, where one station uses either the destination address field or the source address field as the basis for removing a portion of the message.
Abstract: A plurality of stations are used to strip expendable messages from a serial ring communication network. One station uses either the destination address field or the source address field as the basis for removing a portion of the message. Another station uses the status of a control indicia as the basis for removing the remaining portion of the message from the ring. By distributing the stripping function, between multiple stations on the ring, a significant reduction in ring latency is achieved.
TL;DR: A ring latency timer provides a station attached to a token ring network with the capability of obtaining an accurate latency measurement of the ring to which it is attached as discussed by the authors, which can be read via the processor control bus interface.
Abstract: A ring latency timer provides a station attached to a token ring network with the capability of obtaining an accurate latency measurement of the ring to which it is attached. An internal hardware register, which may be read via the processor control bus interface, contains the latest ring latency measurement. A latency interrupt bit, when cleared, enables the latency measurement function. A subsequent interrupt which causes the latency interrupt bit to be set by the chip signals the completion of the latency measurement and the function is once again disabled. The latency register holds the latency information until the interrupt bit is cleared by the processor.