Multiplex communications – Pathfinding or routing – Switching a message which includes an address header
Reexamination Certificate
1999-06-03
2002-08-06
Chin, Wellington (Department: 2664)
Multiplex communications
Pathfinding or routing
Switching a message which includes an address header
C370S412000, C370S466000, C370S474000
Reexamination Certificate
active
06430187
ABSTRACT:
CROSS REFERENCE TO RELATED APPLICATIONS
Not Applicable
STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH OR DEVELOPMENT
Not Applicable
BACKGROUND OF THE INVENTION
The present invention relates generally to the field of data networks.
With the increasing use of Asynchronous Mode Transport (ATM) networks as high capacity, wide-area data transport networks, there has been an increased use of encapsulation and similar techniques to enable ATM networks to provide legacy services. For example, protocols such as the ATM Forum's Local Area Network (LAN) emulation standard are used to emulate a LAN over an ATM network. The emulated LAN uses well-established LAN protocols, such as the well-known Ethernet protocols, and is therefore capable of providing LAN services to a variety of existing customer equipment.
Typically, the ATM network and the legacy services or networks have fundamentally different data transport mechanisms. In ATM networks, data is transported in fixed-length units of 53 bytes called “cells”. In contrast, many other networks employ variable-length data units, referred to herein as “frames”. One basic process performed in equipment that provides frame-based services over an ATM network is the conversion between frames and corresponding groups of cells. The process of converting a frame to a group of cells is known as “segmentation”, and the reverse process as “reassembly”. Logic and/or software that performs the “segmentation and reassembly” or SAR function is ubiquitous in ATM network interface equipment.
The SAR function requires the use of segmentation buffers to temporarily store cells that have been created from received frames, before the cells are transmitted on the ATM network. The segmentation buffers are useful in the segmentation function per se, and they are useful for traffic management purposes as well. The segmentation buffers can absorb cells created from bursts of received frames during periods of reduced transmission capacity on the ATM network. The need to buffer cells can arise from situational circumstances, such as network congestion caused by temporary excess demand, or it may arise due to more deliberate actions, such as the explicit allocation of different priorities to different types of traffic or different customers. During periods of relatively heavy high-priority traffic, the cells of lower-priority traffic will tend to accumulate in the segmentation buffers. After some point, the buffers may become full, resulting in the undesirable discarding of low-priority frames. The threshold point at which frames from a given sender are discarded may be undesirably low if segmentation buffer space is too small, or if one or more high-volume data streams use more than their fair share of the space. In such a case, the quality of service experienced by customers whose frames are discarded can deteriorate to an unacceptable level.
Another feature of ATM networks is the existence of different classes of service. Network users can specify characteristics of their data traffic, and can subscribe for certain levels of network service. For example, a user may desire so-called Constant Bit Rate (CBR) service, which can be useful for real-time data such as video data. CBR service requirements in an ATM network are more stringent than other service classes, such as Available Bit Rate (ABR) or Unspecified Bit Rate (UBR).
In particular, the network attempts to provide different amounts of transmission bandwidth to customers based on subscribed service class. Network devices actively manage the transmission of data cells to achieve the desired allocation of bandwidth, both on a peak basis and on an average basis. It is possible in existing systems that high-volume users may use more than their fair share of transmission bandwidth, causing the quality of service experienced by other users to be degraded.
BRIEF SUMMARY OF THE INVENTION
In accordance with the present invention, techniques are disclosed in which the allocation of buffer space and bandwidth among different network customers connected to a network access device can be more explicitly controlled, enabling improved control over the quality of service experienced by each customer. Additionally, the disclosed techniques enable multiple logical entities, such as LAN bridges, to share a single set of physical resources, saving power, space and cost in network equipment.
According to one disclosed technique, a cell buffer memory is used with segmentation and reassembly (SAR) logic between multiple local area network (LAN) interface logic blocks, on the one hand, and logic interfacing with a cell-based network such as ATM, on the other hand. The cell buffer memory can be programmably partitioned into distinct buffer regions, each buffer region being associated with a respective LAN and with a respective set of virtual connections defined between the SAR logic and the cell-based-network interface logic. The SAR logic segments frames received from each LAN interface logic block into corresponding groups of cells, stores each group of cells into the buffer region associated with the respective LAN, and transfers the groups of cells from each buffer region to the cell-based-network interface logic via the virtual connections associated with the buffer region. The SAR logic and the cell memory are efficiently shared among multiple LAN bridges implemented by the LAN interface logic blocks and the cell-based-network interface logic. Traffic from the different bridges is kept separate, avoiding problems of security and diminished service quality.
According to a disclosed network traffic scheduling technique, the use of allocated transmission bandwidth by the different bridges is actively managed. The technique employs a transmit scheduling table whose entries contain an index value identifying a respective bridge. Each entry represents a cell send opportunity for the respective bridge. A schedule index is continually stepped through the table entries to select connections on which cells are to sent in accordance with a priority scheme reflected in the structure of the table.
Other aspects, features, and advantages of the present invention are disclosed in the detailed description that follows.
REFERENCES:
patent: 5793747 (1998-08-01), Richard Kline
patent: 5818842 (1998-10-01), Burwell et al.
patent: 6157614 (2000-12-01), Pasternak et al.
patent: 6249528 (2001-06-01), Kothary
patent: 0 847 168 (1998-10-01), None
Toshiba Corporation “TC35854F—ATM 155Mb/s SAR Chip” Preliminary Functional Specification, pp. 0-20, 34-35, 41, 83-91, 107-109, 119-120, 164-165 Feb. 7, 1998.
Toshiba Corporation “TC358541F—ATM 155Mb/s SAR Chip”Preliminary Functional Specification, pp. 34-35, 90, 112-113, 124-125, 185-189, Mar. 18, 1999.
Toshiba Corporation “TC358541F—ATM 155Mb/s SAR Chip”Preliminary Functional Specification, pp. 0-20, 34-35, 41, 83-91, 107-109, 119-120, 164-165 Feb. 27, 1998.
Chin Wellington
Fujitsu Network Communications, Inc.
Tran Maikhanh
Weingarten Schurgin, Gagnebin & Lebovici LLP
LandOfFree
Partitioning of shared resources among closed user groups in... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Partitioning of shared resources among closed user groups in..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Partitioning of shared resources among closed user groups in... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2911056