Fujitsu Network Communications, Inc.

Fujitsu Corporation

David N Peck

Thomas A Manning

Robert H Constantin

Stephen R Veit

David E Hammond

Stephen A Caldara

Stephen A Hauser

Mark A W Stewart

Peter Hutzli

Andrew Michal

Salma Abu-Ayyash

A scalable protection method for connection oriented networks includes a source, a destination, and primary nodes interconnected by working path segments between the source and destination. A point-to-multipoint connection with multiple destinations and additional primary nodes is also supported. Each of a number of backup nodes is interconnected with an associated primary node by pre-provisioned shunt segments, and the backup nodes are interconnected among themselves by pre-provisioned protection path segments. Upon occurrence of a failure, the primary node upstream of the failure directs input traffic to a shunt segment, and the backup node associated with the upstream primary node directs traffic from the shunt segment to an output protection segment. Downstream of the failure, a backup node directs traffic from an input protection segment to a shunt segment, and the primary node associated with the backup node directs the traffic from the shunt segment to a downstream working segment. In this manner, traffic is routed around failures via local switching operations.

Scaleable line-based protection for connection oriented communications protocols

A system for allocating bandwidth in a network element. The system determines whether a bandwidth allocation request is associated with a delay sensitive connection type. If so, the system selects a bandwidth allocation algorithm from a first set of bandwidth allocation algorithms. Otherwise, the system selects a bandwidth allocation algorithm from a second set of bandwidth allocation algorithms. The size of the bandwidth allocation request may be examined when selecting which bandwidth allocation algorithm to use. The system may round up the size of the bandwidth allocation request to a predetermined value. The selected bandwidth allocation algorithm may reserve slots in a bandwidth allocation table. Reserved slots that are equally spaced in the bandwidth allocation table avoid undue delays in servicing an associated connection. A binary tree representation of the table is described, wherein leaves of the binary tree correspond to individual table slots. Specific bandwidth allocation algorithms are described which are advantageously applicable when allocating bandwidth to delay sensitive or relatively delay insensitive connection types.

Crosspoint switch bandwidth allocation management

Each switch along a call setup path in a network decides whether the call request should be upgraded from a non-flow-controlled service type to a flow-controlled service type. Each switch bases the decision on whether a boolean variable for the calling subscriber that appears in a management information base (MIB) is set to a value indicating that the subscriber is entitled to the upgrade. In addition, the upgrade is performed only if an interface on the route to the called party on which the call setup request is to be forwarded supports flow-controlled connections. When the call is to be upgraded, the switch replaces the service type indicator in the setup message with a value indicating a flow-controlled service type, and forwards the setup message toward the called subscriber. If the MIB variable is set to a value indicating that the upgrade is not to be performed, it is still possible for a subscriber to request flow-controlled service directly, by indicating this service type in the original call setup message. Each switch also makes a call downgrade decision. When a call setup message is received requesting flow-controlled service but destined for an interface that does not support flow-controlled operation, the switch replaces the service type indicator in the outgoing setup message with a value indicating a non-flow-controlled service type.

Distributed control of service category request mapping

A network device operates to control the clumping of cells on a connection using a rate-based flow-control protocol. In each of successive intervals, a number N is calculated representing the maximum number of cells permitted to be transmitted consecutively. N is calculated as the quotient from dividing a cell delay variation tolerance (CDVT) by the difference between the reciprocal of an allowed cell rate (ACR) and the reciprocal of a link cell rate (LCR). If the calculated quotient is greater than a number N for the connection indicating the maximum number of data cells allowed between forward flow-control cells, then N is set equal to N . A value C is calculated as the largest integer factor of N that is no greater than N, and the interval is then divided into N /C sub-intervals. No more than C cells are allowed to be transmitted on the connection in each of the sub-intervals. This limit is enforced using a credit mechanism in which no more than C credits are extended to a cell transmitter in each sub-interval.

Cell pacing on a network link employing a rate-based flow control protocol with underlying credit-based flow control mechanisms

Network interface logic includes first logic executing a credit-based flow-control protocol (e.g., Quantum Flow Control or QFC) and second logic executing a rate-based flow control protocol (e.g. Explicit Rate or ER). The second logic transfers data cells between a network link and the first logic and inserts and extracts forward (FRM) and backward (BRM) flow-control/resource management (RM) cells. The second logic further (i) obtains cell buffer utilization information from the first logic, (ii) supplies transmission credits to the first logic based on rate information in BRM cells, (iii) induces the first logic to emit &ldquo;dummy&rdquo; RM cells when an RM cell is to be transmitted, and (iv) converts the dummy cells to either FRM or DRM cells for forwarding on the link. Rate information in the transmitted RM cells is calculated from the buffer utilization information and rate information in extracted BRM cells.

Network switch supporting rate-based and credit-based flow control mechanisms on a link-by-link basis

A system for allocating resources at network elements along a path through a communications network. Resources are allocated at a network element along the path in response to a connection request message. The delay budget is the minimum of a calculated cell transfer delay and a calculated cell delay variation. The calculated cell transfer delay is determined by finding a difference between a target end to end cell delay and an accumulated cell delay, and then dividing that difference by a number of network elements remaining in the path between the network element and the end of the path. To determine the calculated cell delay variation, the network element first finds a difference between the Max CDV QoS parameter and the sum of the accumulated cell delay variation and the switching delay for the network element. The calculated cell delay variation is then equal to that difference divided by the number of network elements remaining in the path between the network element and the end of the path. The target end to end cell delay is for example determined by the first network element in the path as equal to the Max CTD QoS parameter minus the end to end link propagation delay for the path and the link propagation delay for the incoming link to the first network element over which the message was received. The first network element generates and forwards a modified connection request message including the determined target end to end cell delay value, as well as a designated transit list indicating each network element in the prospective path, and including a current network element pointer.

Network element delay allocation for efficient use of network resources

The invention comprises a method and apparatus for discarding frames in a communications device. In accordance with the method of the invention, a plurality of cells are received representing a plurality of frames wherein each cell comprises a data portion and a header portion including a cell loss priority indicator. At least some of the cells are stored in a buffer. It is determined if the number of cells in the buffer exceeds a first threshold when a first cell including an end of file marker is received. A series of cells received between the first cell and a second cell are discarded if the first threshold was exceeded in the determining step and if the cell loss priority indicator for a cell in the series of cells is set to a first state. The second cell comprises a cell including an end of file marker.

Growing community of inventors

Northbridge, MA, United States of America

David N Peck