CN1783828A - Space division exchanging device and information element exchanging method - Google Patents
Space division exchanging device and information element exchanging method Download PDFInfo
- Publication number
- CN1783828A CN1783828A CNA2004101009029A CN200410100902A CN1783828A CN 1783828 A CN1783828 A CN 1783828A CN A2004101009029 A CNA2004101009029 A CN A2004101009029A CN 200410100902 A CN200410100902 A CN 200410100902A CN 1783828 A CN1783828 A CN 1783828A
- Authority
- CN
- China
- Prior art keywords
- cell
- switching
- module
- equipment
- input
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Landscapes
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
This invention relates to a space division equipment and cell exchange method. It contains obtaining the information of destination terminal from cell head to be exchanged, generating arbitration result according to relative information, said result containing exchange information of each input terminal cell, scheduling relative cell to input terminal of exchange equipment, controlling exchange equipment to exchange input terminal cell to specified export terminal. Said space division equipment includes switching matrix module, input process module, arbitration module, exchange distribution module and output process module. Said invention can realizes ultra large capacity exchange network with fine expandability without large capacity of cell buffer memory.
Description
Technical field
The present invention relates to data communication field, relate in particular to a kind of space switching equipment and cell switching method with arbitration.
Background technology
Switching network (Switch Fabric) is the nucleus module in the router, is used to finish between a plurality of ports the exchange of bag or cell, promptly the cell switching that arrives input port to corresponding output port.As shown in Figure 1, be a more typical router topology, the function of switching network is finished jointly by the switching equipment (FabricDevice) that is positioned at the queue management equipment (Queue Management) on the ply-yarn drill (Line Card) and exchange on the network interface card (Fabric Card).On Way in, the cell of coming in from this ply-yarn drill is carried out virtual output work queue, and (Virtual Output Queuing VOQ), and delivers to corresponding exchange equipment early and gets on to finish exchange; On down direction, queue management equipment is ranked by output port to cell once more, guarantees that cell is guaranteeing to send to the corresponding port under the suitable Quality of Service prerequisite.One or more switching equipment (Fabric Device) on the exchange network interface card are finished the function of exchange of flow between each ply-yarn drill, promptly are responsible for effectively the flow of each ply-yarn drill being exchanged on the purpose ply-yarn drill.
Fig. 2 is a kind of switching equipment that adopts the shared buffer memory structure at present, when work, be forwarded to all cells on the switching equipment as soon as possible by ply-yarn drill, these cells of buffer memory on switching equipment, and all cells are ranked by port and priority, by scheduler these formations are dispatched, guaranteeing under the prerequisite of Quality of Service cell switching to the purpose ply-yarn drill.
Adopt the shortcoming of this scheme to be:, need the shared buffer memory of larger capacity, and required buffer memory RAM bandwidth to be very big owing to adopt the shared buffer memory exchange of concentrating.For example, a shared buffer memory switching equipment of supporting 16 10Gbps ports because each cell all need write shared buffer memory, also need be read shared buffer memory, therefore required readwrite bandwidth is at least: 16 * 10Gbps * 2=320Gbps, this is to be difficult to reach with present main flow semiconductor technology.This just means that also this structure of employing is difficult to capacity is done greatly.In addition, the shared buffer memory of adopt concentrating, need be in chip integrated jumbo RAM, just make that also the production cost of this equipment and power consumption are all very high.
Summary of the invention
The objective of the invention is at the deficiencies in the prior art, a kind of space switching equipment and cell switching method are provided, big to exist cell buffer memory capacity demand in the solution prior art, the problem that power consumption and cost are high.
For addressing the above problem, the invention provides following technical scheme:
A kind of method of cell-switching, by switching equipment with the cell switching of a plurality of input ports to corresponding output port; It is characterized in that this method comprises the steps:
From the header of respectively treating cell-switching, obtain the relevant information of the destination interface that need exchange to about cell;
Produce arbitration result according to each relevant information, this arbitration result comprises the information of each input port cell switching to the appointed output terminal mouth;
Dispatch the input port of corresponding cell according to each arbitration result to described switching equipment; With
Utilize arbitration result control switching equipment with the cell switching of input port output port to appointment.
According to said method:
Produce one group of arbitration result in cycle at a cell, this arbitration result is finished cell switching through being used to control described switching equipment after the time-delay in one or more cell cycle.
Described arbitration result is inserted in the field of header and sent back to the individual queue management equipment, queue management equipment is according to the arbitration result scheduling cell of the header echo port to switching equipment.
In the header that sends back to the individual queue management equipment, also add synchronous time mark, queue management equipment is attached to this synchronous time mark according to the arbitration result in this header and is dispatched in the cell of switching equipment, and switching equipment realizes that according to the synchronous time mark in each input port cell cell is synchronous.
Before cell-switching, synchronous time mark in the header of this cell and the current pairing markers of arbitration result that is used for controlling switching equipment are compared, and when both do not conform to the report configuration error; And/or the destination interface in destination interface in the header and the current arbitration result that is used for controlling switching equipment compared, and when both do not conform to the report configuration error.
A kind of space switching equipment, this equipment comprises:
Switch fabric module has a plurality of inputs and corresponding output, the extremely corresponding output of cell switching that this module is come in each input;
A plurality of input processing modules, the output of each module are connected with an input of switching matrix respectively, and this module is extracted header information from the cell that receives, and cell is outputed to switching matrix synchronously;
Arbitration modules is connected with each input processing module, produces the arbitration result that is used for cell-switching according to the solicited message in the header information;
The exchange configuration module is connected with switch fabric module with arbitration modules, and the arbitration result that arbitration modules is exported disposes described switch fabric module and carries out cell switching;
A plurality of output processing modules, each module respectively with output of switch fabric module be connected with the moderator module respectively.
Described space switching equipment also comprises: the route inspection module, be connected with described exchange configuration module, each input processing module and each output processing module, and be used for each cell that is about to exchange is carried out route inspection and validity check.
A kind of switching system comprises the up administration queue and down management formation and a plurality of aforesaid space switching equipment that are used to dispatch cell; Described up administration queue and down management formation are a plurality of, each up administration queue is connected with a input processing module in each space switching equipment respectively, and each down management formation is connected with a output processing module in each space switching equipment respectively.
The present invention has following beneficial effect:
1, the present invention adopts a big space switching matrix to realize cell switching, does not need jumbo cell buffer memory when carrying out cell switching, need not rank in equipment, makes that realization is very simple, easily the port number of monolithic and capacity is done greatly.
When 2, adopting the core that space switching equipment of the present invention in return nets, can adopt one or more this switching equipment flexibly,, realize the switching system of various different capabilities in conjunction with the queue management equipment on the ply-yarn drill; When needs are expanded, can realize that therefore, the flexibility and the autgmentability of its networking are very good by number that increases space switching equipment and the port number that increases space switching equipment.
3, being accomplished to arbitration result from arbitration among the present invention, to be used to dispose the time-delay of switching matrix adjustable, thereby can be asynchronous fully between the individual queue management equipment, make system not high, make system design become easy the delay requirement between space switching equipment and the queue management equipment.
Description of drawings
Fig. 1 is a typical router topology schematic diagram in the prior art;
Fig. 2 is the structural representation of the switching equipment of shared buffer memory in the prior art;
Fig. 3 is the structural representation of band space switching equipment of the present invention;
Fig. 4 is the structural representation of switching matrix of the present invention;
Fig. 5 is the structural representation that the present invention exchanges configuration module;
Fig. 6 is the structural representation of the exchange configuration module of replacement scheme of the present invention;
Fig. 7 is the system applies schematic diagram of space switching equipment of the present invention.
Embodiment
As shown in Figure 3, space switching equipment of the present invention comprises: switch fabric module SMX, a plurality of input processing module IP, arbitration modules ARB, exchange configuration module CFG, route detection module RCK, a plurality of output processing module OP.Wherein:
Switch fabric module SMX with a plurality of inputs and corresponding output is a choke free space switching matrix, and the cell switching that this matrix is used for each input is come in is to corresponding output;
Each input processing module IP that its output is connected with an input of switching matrix respectively is used for extracting header information from the cell that receives, and cell is outputed to switching matrix;
The arbitration modules ARB that is connected with each input processing module can produce the arbitration result that is used for cell-switching according to the solicited message in the header information;
With the exchange configuration module CFG that arbitration modules is connected with switch fabric module, be used for the arbitration result of arbitration modules output is converted to the configuration information of switching matrix, and buffer memory in addition;
With exchange configuration module, each input processing module and the route inspection module that each output processing module is connected respectively, be used for each cell that is about to exchange is carried out route inspection and validity check;
With cell information after output of switch fabric module and each output processing module OP that arbitration modules is connected respectively are used for exporting exchange and arbitration result is attached to header sends.
As shown in Figure 4, described switch fabric module comprises:
A plurality of selectors with multichannel input and single channel output, one of them input of each selector links to each other with an input processing module, and this input also is connected with an input of other selectors respectively.
In addition, present embodiment also provides two kinds of structures that can supply the exchange configuration module of replacement:
As shown in Figure 5, the exchange configuration module comprises:
Plurality of memory cells (i.e. 8 configuration information modules among the figure) is used for the cache arbitration result;
The arbitration result that each cell is arbitrated out converts the required form of switching matrix configuration information (selection signal) to, and deposits it in 8 configuration information pieces by selector 1 by the LOCAL_SYNC_TS (synchronous time mark) in current cell cycle; When sense information is used to dispose switching matrix, again by selector 2 according to LOCAL_SYNC_TS-q (time-delay) choose a configuration set information, be used to dispose switching matrix.
As shown in Figure 6, the exchange configuration module comprises:
A plurality of shift registers (in this example be 8, but be not limited to 8) are used for the arbitration result that cache arbitration device module is exported;
Have the selector of multichannel input and single channel output, each shift register is connected with input of selector respectively, and this selector is selected the information output of a register from described a plurality of shift registers.
Among Fig. 6, the arbitration result of coming out from moderator fixedly deposits shift register CFG0 in, later each cell cycle, the information of each arbitration result group realizes displacement, be that CFG0 moves into CFG1, CFG1 moves into CFG2 ..., be sent to the configuration information of switching matrix and then select by a selector, the selection signal of making selector with signal q gets final product here.
Arbitration result converts the switching matrix configuration information to can just use all the other module converts before selected cell, also can finish in selected cell, or all can at the interface of memory cell.
Above-mentioned space switching equipment can be finished the cell space switching of n port to n port.In order to state conveniently, follow-up explanation hypothesis n is 16, and in fact, n can be for than 16 much bigger numbers, as 32,64, and 128 etc., can both under present semiconductor technology, realize.Below the process of cell-switching is described.
The cell of sending into space switching equipment is a fixed length, except payload (Payload), also comprises the information that some are finished the required information of space switching and need transparent transmission.These information are formed a header.Suppose that cell length is a L byte, so, if an input port bandwidth is the B byte per second, then switching equipment tradable every port cell number in 1 second is B/L, and we claim that L/B is a cell cycle, and promptly switching equipment is finished the time of one group of cell switching.During this period of time, switching equipment is finished the exchange of one group of cell (one of each port).
1, from the header of respectively treating cell-switching, obtains the relevant information of the destination interface that need exchange to about cell by input processing module IP.
The individual queue management equipment of joining with space switching equipment is before by space switching devices exchange cell, cell switching be informed space switching equipment to hope to the port information of destination interface, this realizes by band solicited message REQ in header.Solicited message REQ is the information of a n bit, and the destination interface of the corresponding space switching equipment of wherein every bit indicates whether to have cell to be sent to this destination interface, for " 1 " representative has cell, is the no cell of " 0 " representative.This n bit can have any position to be " 1 ".Like this, space switching equipment just can be received the solicited message of n n bit.Therefore, the IP module in the space switching needs at first the solicited message REQ in the header of corresponding port cell flow to be sampled.Because the n that comes in from each port group solicited message is not need synchronously, only needs directly the request of up to the present receiving to be sampled, as the foundation of next step arbitration at each cell certain fixed time in the cycle.
2, arbitration modules produces arbitration result according to each relevant information.
The selector that is used to control switching matrix selects signal SEL to obtain by arbitration, the solicited message that arbitration modules obtains according to the IP module samples, determine which input port each output port allows send information toward this output port, thereby obtain to comprise the arbitration result of selecting signal.Because the solicited message of all carrying the n bit in the header that each input port of space switching matrix is come in, like this can be so that moderator is considered all queue management equipment that switching requirement is arranged simultaneously, and make the utilance of space switching matrix reach the highest.Produce one group of arbitration result in cycle at each cell
3, this arbitration result is sent back to individual queue management equipment scheduling cell, simultaneously arbitration result is cached in the exchange configuration module.
Arbitration result need send back to queue management equipment, as the foundation of queue management equipment scheduling cell.This allows field (GRANT) to realize that GRANT indicates corresponding queue management equipment by add one at header, shows whether can send cell and cell sent to which space switching device port; The figure place of GRANT field is log2n+1, and n=16 in the present embodiment then needs 5, and wherein whether a bit representation can send cell, and all the other 4 bit representations can send which port in cell to 16 port.
The arbitration result that produces in cycle at a cell can not be used to dispose switching matrix immediately, needs the fixing q cycle of time-delay, and this is to send into the IP module because equipment need be waited for cell-switching from external equipment, could dispose switching matrix.The delay cycle q here can join in chip, and when system design, after can sending according to an arbitration result, the actual time delay between the cell of sending here to outside this arbitration result of queue management device responds disposes.In real system, this time-delay of each bar link is devious, suppose that wherein maximum time-delay is p cell cycle, then q should be configured to the value more slightly bigger than p, because it is if littler than p, treat that then cell-switching also do not send into the IP module from external equipment,, then make the exchange time-delay of cell become too big if bigger too much than p.
Because arbitration result will just be used to dispose switching matrix more after clap, therefore, earlier these arbitration result buffer memorys to be got up, this realizes by exchange configurator CFG.CFG finishes the transfer process of switching matrix configuration information when finishing the buffer memory step.
Because the SEL of n selector selects signal to switch simultaneously in the switching matrix, this just means that cell will carry out synchronously before being sent to the switching matrix exchange, each GRANT signal that space switching equipment was sent toward n queue management equipment in the same cell cycle transmits by cell, it is different to be sent to the link that the cell of different queue management equipment walks, this just causes different queue management equipment to receive the asynchronism(-nization) of the cell that contains this GRANT information, and common individual queue management equipment is on different ply-yarn drills, be fully asynchronous between them, therefore, to arrive the time of space switching equipment be different to the cell of sending according to this GRANT information.For this reason, need the cell that a kind of synchronization mechanism guarantees a collection of GRANT and send into the switching matrix exchange at one time.
This synchronization mechanism is by when sending GRANT information, also synchronous time mark SYNC_TS[2:0 of affix in header] territory, this territory is one 3 a field, add 1 in each cell cycle, be referred to as LOCAL_SYNC_TS, each cell cycle is in the header of all queue management equipment transmissions, with synchronous time mark be the same, all be LOCAL_SYNC_TS, like this, after queue management equipment receives a cell, GRANT territory and SYNC_TS territory are taken off simultaneously, again this SYNC_TS value is attached to according to this GRANT information to go out after the team in the header that sends toward space switching equipment to give space switching equipment, like this, space switching equipment just can be realized according to the SYNC_TS territory that each port is come in the cell synchronously.
In input processing module IP, finishing synchronously before the cell switching, each IP module all has the buffer memory (BUFFER) of 4 cell degree of depth, and this BUFFER realizes by the cell that enters the IP module is carried out buffer memory synchronously.This BUFFER is divided into 4 pieces, can store a cell for every, and cell selects to write which piece, just SYNC_TS[1:0 according to low two of SYNC_TS in its header] as the highest two bit address of BUFFER.It is synchronous that the BUFFER of all IP modules reads, promptly the address of reading of same bat BUFFER is the same in all IP modules, the highest two of this address is LOCAL_SYNC_TS-q, as above-described, the q here is the value that arbitration result of an adjustable expression is used to dispose switching matrix at what all after dates.
4, before cell-switching, by route detection module RCK the cell that respectively is about to exchange is done group and check (validity check) and route inspection.The synchronous time mark that is about in the header of this cell compares with the current pairing markers of arbitration result that is used for disposing switching matrix, and reports configuration error when both do not conform to; Destination interface in destination interface in the header and the current arbitration result that is used for disposing switching matrix is compared, and when both do not conform to the report configuration error.Specifically comprise:
(1) group is checked: as previously mentioned, arbitration result be q cell week after date just be used for disposing switching matrix, for the ease of system debug, avoid because the configuration error of q causes the cell switching mistake, the RCK module is done the group inspection to the module that is about to exchange.
In order to reduce logic as far as possible, the synchronous time mark that present embodiment is attached in the cell is 3, but in input processing module, only used the BUFFER of four cell degree of depth, only be used for selecting to deposit in which cache blocks with low 2 in the synchronous time mark, rather than with the BUFFER of 8 cell degree of depth, this is because the difference that minimum arbitration result → GRANT extractions → cell goes out between whole loop cycle min_loop_slot of team and the maximum loop cycle max_loop_slot generally can not surpass 4 cell cycles (generally all being easy to accomplish below 3 even 2 cell cycles).But a kind ofly may be, then may cause sending into the SYNC_TS[2 in the cell of switching matrix if q is not provided with correctly] with the SYNC_TS[2 of configuring matrix information accordingly] different, cause exchanging mistake.In order to check out and avoid this situation, the CFG module has not only been preserved the switching matrix configuration information in the configuration information registers group, simultaneously also SYNC_TS[2:0] preserve, like this, in every cell cycle, when from 4 buffer memorys, selecting a cell and being sent to switching matrix and exchanging, simultaneously can be with the SYNC_TS[2 in this header] with the pairing SYNC_TS[2 of configuration information that is about to be used for disposing switching matrix] compare, be referred to as group inspection, if check and do not pass through, represent that promptly the group number (being the bit2 of SYNC_TS) of the cell-switching and the actual exchange cell of current expectation is not inconsistent.Causing organizing the reason that inspection do not pass through generally is because q is provided with excessive or too small causing.This just provides reference for system debug, and has guaranteed the reliability of exchange.
(2) route inspection: reliable for what guarantee to exchange, promptly guarantee current switching matrix configuration information and treat that the destination interface that cell-switching need be routed to is consistent, a route inspection (the abbreviation RCHK of Routing Check) territory is arranged in header, be used to represent the destination interface of this cell.The RCK module compares this destination interface and present configuration information when each cell of exchange, to guarantee cell switching to correct destination interface, being about to destination interface in the header and the destination interface in the current arbitration result that is used for disposing switching matrix compares, and when both do not conform to, report configuration error information, and cause interruption.
5, utilize arbitration result configuration switch fabric module, make it the cell switching of input port output port to appointment.
The exchange of cell is finished in switch fabric module, and switching matrix can be regarded the selector that n n road is input to one tunnel output as.As shown in Figure 4, these selectors select signal SEL to select this road output from which input to select according to one, like this, in the cycle, can select signal to finish the space switching of cell at each cell by switching this.For simplicity, switching matrix is synchronous to the exchange of 16 port cells, and promptly the selection signal of all n selector switches simultaneously.
More than describe the switching method of space switching equipment and cell in detail, in actual applications, the use that can walk abreast of a plurality of switching equipment, the present invention provides a kind of switching system in view of the above, this system comprises a plurality of up administration queue and down management formation, and is connected in a plurality of space switching equipment therebetween.As shown in Figure 7, used m switching equipment on exchange network interface card (Fabric Card), total like this exchange capacity has just increased m doubly.Upstream queue management equipment on the ply-yarn drill (Ingress Queue Device) is distributed on a plurality of switching equipment any to cell, exchanges to down queue management equipment (Egress QueueDevice) by switching equipment.Upstream queue management equipment here and down queue management equipment can be two equipment, also can be to finish in an equipment.
Use this structure, just realized mutual backup between a plurality of space switching equipment, after any space switching equipment broke down, the cell of any upstream queue management equipment still can route to any down queue management equipment by other space switching equipment.
In addition, this structure also makes the capacity extension of switching network be easy to, and by increasing space switching equipment, just can realize the expansion of capacity.
Obviously, those skilled in the art can carry out various changes and modification to the present invention and not break away from the spirit and scope of the present invention.Like this, if these are revised and modification belongs within the scope of claim of the present invention and equivalent technologies thereof, then the present invention also is intended to comprise these changes and modification interior.
Claims (11)
1, a kind of method of cell-switching, by switching equipment with the cell switching of a plurality of input ports to corresponding output port; It is characterized in that this method comprises the steps:
From the header of respectively treating cell-switching, obtain the relevant information of the destination interface that need exchange to about cell;
Produce arbitration result according to each relevant information, this arbitration result comprises the information of each input port cell switching to the appointed output terminal mouth;
Dispatch the input port of corresponding cell according to each arbitration result to described switching equipment; With
Utilize arbitration result control switching equipment with the cell switching of input port output port to appointment.
2, the method for claim 1 is characterized in that, produces one group of arbitration result in the cycle at a cell, and this arbitration result is finished cell switching through being used to control described switching equipment after the time-delay in one or more cell cycle.
3, method as claimed in claim 1 or 2 is characterized in that, described arbitration result is inserted in the field of header and is sent back to the individual queue management equipment, and queue management equipment is according to the arbitration result scheduling cell of the header echo port to switching equipment.
4, method as claimed in claim 3, it is characterized in that, in the header that sends back to the individual queue management equipment, also add synchronous time mark, queue management equipment is attached to this synchronous time mark according to the arbitration result in this header and is dispatched in the cell of switching equipment, and switching equipment realizes that according to the synchronous time mark in each input port cell cell is synchronous.
5, method as claimed in claim 4, it is characterized in that, before cell-switching, synchronous time mark in the header of this cell and the current pairing markers of arbitration result that is used for controlling switching equipment are compared, and when both do not conform to the report configuration error; And/or the destination interface in destination interface in the header and the current arbitration result that is used for controlling switching equipment compared, and when both do not conform to the report configuration error.
6, a kind of space switching equipment is characterized in that comprising:
Switch fabric module has a plurality of inputs and corresponding output, the extremely corresponding output of cell switching that this module is come in each input;
A plurality of input processing modules, the output of each module are connected with an input of switching matrix respectively, and this module is extracted header information from the cell that receives, and cell is outputed to switching matrix synchronously;
Arbitration modules is connected with each input processing module, produces the arbitration result that is used for cell-switching according to the solicited message in the header information;
The exchange configuration module is connected with switch fabric module with arbitration modules, and the arbitration result that arbitration modules is exported disposes described switch fabric module to carry out cell switching;
A plurality of output processing modules, each module respectively with output of switch fabric module be connected with the moderator module respectively.
7, space switching equipment as claimed in claim 6 is characterized in that also comprising:
The route inspection module is connected with described exchange configuration module, each input processing module and each output processing module, is used for each cell that is about to exchange is carried out route inspection and validity check.
8, as claim 6 or 7 described space switching equipment, it is characterized in that described switch fabric module comprises:
A plurality of selectors with multichannel input and single channel output, one of them input of each selector links to each other with an input processing module, and this input also is connected with an input of other selectors respectively.
9, space switching equipment as claimed in claim 8 is characterized in that, described exchange configuration module comprises:
Plurality of memory cells is used for the memory arbitration result;
First selector has one tunnel input and multichannel output, the arbitration result that is used for selecting the described arbitration modules of a cell stores of plurality of memory cells to export;
Second selector has multichannel input and one tunnel output, is used for selecting a memory cell output arbitration result from described plurality of memory cells;
Perhaps, described exchange configuration module comprises:
A plurality of shift registers are used for the arbitration result that cache arbitration device module is exported;
Have the selector of multichannel input and single channel output, each shift register is connected with input of selector respectively, and this selector is selected the information output of a register from described a plurality of shift registers.
10, a kind of switching system comprises the up administration queue and the down management formation that are used to dispatch cell; It is characterized in that this switching system also comprises a plurality of space switching equipment as claimed in claim 6; Described up administration queue and down management formation are a plurality of, each up administration queue is connected with a input processing module in each space switching equipment respectively, and each down management formation is connected with a output processing module in each space switching equipment respectively.
11, switching system as claimed in claim 10 is characterized in that, described upstream queue management equipment and down queue management equipment are separate equipment, perhaps is same equipment.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2004101009029A CN1783828B (en) | 2004-12-02 | 2004-12-02 | Space division exchanging device and information element exchanging method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2004101009029A CN1783828B (en) | 2004-12-02 | 2004-12-02 | Space division exchanging device and information element exchanging method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1783828A true CN1783828A (en) | 2006-06-07 |
| CN1783828B CN1783828B (en) | 2010-04-28 |
Family
ID=36773611
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2004101009029A Active CN1783828B (en) | 2004-12-02 | 2004-12-02 | Space division exchanging device and information element exchanging method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN1783828B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101631067B (en) * | 2008-07-14 | 2011-12-21 | 华为技术有限公司 | Method, device and system for packet switching |
| CN108199976A (en) * | 2017-12-28 | 2018-06-22 | 天津芯海创科技有限公司 | Switching equipment, exchange system and the data transmission method for uplink of RapidIO networks |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| GB9828144D0 (en) * | 1998-12-22 | 1999-02-17 | Power X Limited | Data switching apparatus |
| WO2004032437A2 (en) * | 2002-10-02 | 2004-04-15 | Cisco Technology, Inc. | Method and apparatus for processing superframes using an arbitration system |
| CN1689281A (en) * | 2002-10-02 | 2005-10-26 | 皇家飞利浦电子股份有限公司 | Weight adaptation in packet switches |
-
2004
- 2004-12-02 CN CN2004101009029A patent/CN1783828B/en active Active
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101631067B (en) * | 2008-07-14 | 2011-12-21 | 华为技术有限公司 | Method, device and system for packet switching |
| CN108199976A (en) * | 2017-12-28 | 2018-06-22 | 天津芯海创科技有限公司 | Switching equipment, exchange system and the data transmission method for uplink of RapidIO networks |
| CN108199976B (en) * | 2017-12-28 | 2021-05-04 | 天津芯海创科技有限公司 | Switching equipment, switching system and data sending method of RapidIO network |
Also Published As
| Publication number | Publication date |
|---|---|
| CN1783828B (en) | 2010-04-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN1201532C (en) | Quick-circulating port dispatcher for high-volume asynchronous transmission mode exchange | |
| SE461002B (en) | DIGITAL COMMUNICATION SYSTEM WITH DIVIDED CONTROL | |
| US7324537B2 (en) | Switching device with asymmetric port speeds | |
| CN101052013A (en) | Method and system for realizing network equipment internal managing path | |
| CN101189843A (en) | Electronic device and method of communication resource allocation | |
| CN1264531A (en) | Switching fabric arrangement with time stamp function | |
| CN101873253A (en) | Buffered crossbar switch system | |
| CN1859263A (en) | Cell re-ordering method, cell processing method and device using said method and device | |
| CN1050657A (en) | Be used for the irrelevant system of telecommunication switching system and resource | |
| CN1440608A (en) | System having meshed backplane and process for transferring data therethrough | |
| CN1783839A (en) | Flow control credit updates for virtual channels in the advanced switching (as) architecture | |
| JPH08265270A (en) | Transfer line assignment system | |
| CN1357188A (en) | Data switching method and appts. | |
| CN102158403B (en) | Efficient data stream transmission communication structure suitable for on chip network and operating method thereof | |
| CN103530245B (en) | A kind of SRIO interconnecting and switching device based on FPGA | |
| CN1829199A (en) | Switch for integrated telecommunication networks | |
| CN110113271A (en) | A kind of MPI application acceleration system and method based on photoelectricity hybrid switching network | |
| CN100417136C (en) | Down queue fast back pressure transmitting based on three-stage exchange network | |
| CN103257946A (en) | High-speed interconnecting method of controllers of tight-coupling multi-control storage system | |
| CN1109307C (en) | System for interchanging data between data processor units having processors interconnected by common bus | |
| CN102111327A (en) | Method and system for cell dispatching | |
| CN1130951C (en) | Segmented performance monitoring of multi-stage ATM node | |
| CN100401718C (en) | Data frame group broadcasting duplicating method and system | |
| CN100536460C (en) | Device for dispatching and arbitration | |
| CN1466348A (en) | High speed link control protocol transmission processing/module and data processing/method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |