CN101656900A - OPS node optical buffering device and method based on load selection - Google Patents
OPS node optical buffering device and method based on load selection Download PDFInfo
- Publication number
- CN101656900A CN101656900A CN200910190827A CN200910190827A CN101656900A CN 101656900 A CN101656900 A CN 101656900A CN 200910190827 A CN200910190827 A CN 200910190827A CN 200910190827 A CN200910190827 A CN 200910190827A CN 101656900 A CN101656900 A CN 101656900A
- Authority
- CN
- China
- Prior art keywords
- output
- fdl
- optical
- formula
- shared
- 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
The invention asks for protecting optical buffering device and method based on load selection, relating to the technical field of optical fiber communication. Aiming at the characteristics of low optical fiber utilization rate, large volume and high cost of the prior optical packet switching node buffering structure, the invention provides a novel optical packet switching node buffering device based on load selection. When the network service load is calculated to be lower than a threshold value by a switch control module of an OPS node, an output port of the buffering device is communicated to an output port of the OPS node by an optical switch in a switching mode to expand and output FDL buffer capacity; when the network service load is larger than the threshold value, the output port ofthe buffering device is connected to the expansion input port of the OPS node in a switching mode and used as a feedback share buffer, and optical fiber delay lines of the optical buffering device are allowed to enter the multiple-time feedback connection for many times to reduce the problem that the packet loss rate is quickly increased for the limited FDL buffer capability when the high load isreduced. The device and the method better solve the problem of port competition of an optical packet switching network.
Description
Technical field
The present invention relates to fiber optic communication field, be specifically related to a kind of optical buffer device that is used for light packet-switched core network node.
Technical background
Light packet switching OPS (Optical packet switching) is a kind of novel switching technology of rising in recent years, it combines the characteristics of electronic type group technology and light switching technology, support flexibly, the packet level exchange of arbitrarily small granularity, thought the most desirable exchanged form of network light/electricity/optical information bottleneck problem not solve by industry.In the OPS network, when a plurality of groupings of same wavelength need same port output from intermediate node simultaneously, will compete, and very big based on packet burst in the Internet network of IP operation, the then professional probability that generation is competed in OPS network core node is also bigger.Competition solution commonly used mainly contains three kinds, be fibre delay line (FDL) buffer memory, wavelength Conversion and penetrate routing plan partially, they are respectively on the time-domain, solve race problem on the wavelength domain and on the spatial domain, because wavelength shifter prematurity still, only be fit to specific lattice l network and penetrate route partially, therefore, the FDL buffer memory is realized and the most frequently used light territory caching technology at present easily.Insert storage at random because FDL can not provide, the FDL framework and the controlling mechanism of OPS node are still waiting further research, to improve the performance that finite population FDL buffer efficiency and competition solve.
Solve in the light grouping competition structure at existing FDL, FDL is configured in the problem of the poor-performing that is cached with hol blocking and solution competition at OPS input node place; The buffer memory that FDL is configured in node output solves the grouping competition better performances, but exists intra-node need quicken and the bigger problem of demand FDL number; FDL is configured in that the internodal feedback of output one input is shared and the output type shared buffer memory has reduced demand FDL number, but there is repeatedly circulation losses in light signal and reduces the risk of signal to noise ratio.Therefore, study novel FDL configuration structure and control strategy, the present invention is in conjunction with output buffers and share the advantage of feeding back buffer memory, and the competition resolution ability of finite population FDL and FDL utilance are improved, and reducing OPS node volume and improving the OPS performance is that OPS realizes extremely urgent realistic problem.
Summary of the invention
At existing smooth packet switching OPS node configuration fibre delay line FDL buffer memory solve that grouping competition exists that buffer memory granularity bulky, that FDL is discrete makes that its utilance is not high, the OPS defective that the competition resolution ability descends when the professional heavy load, the present invention has designed a kind of device and control method of selecting the shared formula FDL buffer memory of output/feedback based on load.
The present invention designs a kind of OPS node optical buffer storage of selecting based on load, and this device comprises: the output/feedback based on the load selection of optical multiplexer/demodulation multiplexer, optical switch, optical branching device/mixer, the configuration of degeneracy formula is shared formula FDL module, formula output FDL cache module and switching control module are shared in segmentation.Wherein, output/feedback is shared N ports share M root fibre delay line FDL of formula FDL module, the output of OPS node connects optical multiplexer by switching control module, sharing formula FDL module input by optical multiplexer and output/feedback links to each other, output/feedback is shared formula FDL module output and is connected optical switch by fibre delay line, the output 1 of optical switch connects optical demultiplexer, and closes the road by a plurality of output ports of optical combiner and OPS node; The output 2 of optical switch feeds back the input port that connects the OPS node respectively, and whether switching control module surpasses threshold value decision selective light output switching terminal 1 or 2 connections according to arriving traffic load values; The part output of switching control module connects segmentation and shares formula output FDL cache module.
Output/feedback is shared formula FDL module has M input/output end port that D is provided, 2D, 3D, ..., the time delay granularity of MD, M the output port that output/feedback is shared formula FDL module is connected or is connected by the optical combiner of optical demultiplexer with OPS node output port by an expansion M input port of optical switch with the input of OPS node, uses as output capacity buffer memory expansion module or feedback contention-based solution module.
OPS node statistics arrives professional load value, when business load is lower than when threshold value is set, optical switch on the buffer memory output line is communicated with the corresponding output port of OPS node by demodulation multiplexer, and is multiplexing with the shared formula FDL buffer of the segmentation of output configuration, the output buffers capacity of Extended Capabilities Port; When the business load that arrives the OPS node during more than or equal to threshold value, optical switch on the buffer memory output line connects M Extended Capabilities Port of OPS node switching control module input, output/feedback is shared formula FDL module as the use of feedback shared buffer memory device, and the repeatedly circulation that allows to divide into groups enters the feedback cache module and plays better grouping competition solution effect.Simultaneously, the OPS node also disposes N input/output end port, select an optical fiber by switching control module at each output port, every optical fiber is divided into B section (the output port buffer memory degree of depth), every period unit interval postpones to be D, the optical fiber of every section upper level just can be shared all optical fiber of its next stage, has realized that each output port configuration capacity is the shared formula FDL buffer memory of segmentation of BD (being that the buffer memory degree of depth is that B chronomere doubly postpones).
Fibre delay line is shared on the abundant realization of this device single port, and B+M≤N (wherein, B is the output port buffer memory degree of depth, and M is the port number that output/feedback is shared formula FDL module, and N is an OPS node input/output end port number).By contrasting with output FDL buffer memory, segmented of the present invention is shared the needed FDL number of output buffers and is obviously reduced.
Output/feedback that the present invention proposes is shared formula FDL buffer memory according to arriving business load size adjustment mode of operation, and the competition resolution ability of formula buffer memory is shared in the segmentation that increases the OPS node configuration.Its control method is: the OPS node is added up arriving professional load value ρ, when traffic load values is lower than threshold value H, the output port that output/feedback is shared formula FDL buffer storage is communicated with optical demultiplexer by optical switch, be connected with the optical combiner of the output port of OPS node, buffer storage is operated under the expansion output shared buffer memory pattern, as the dilatation module of OPS node output buffers, play the output buffers ability effect that increases node; When business load is increased to threshold value H, OPS node control optical switch switches, each output port feedback that output/feedback is shared formula FDL cache module is connected to OPS node input port, by feedback FDL output of OPS node and input port are communicated with, in the mode of the buffer storage FDL that can be recycled as feedback shared buffer memory module, allow grouping repeatedly to enter the fibre delay line of feedback, the problem that the packet loss that causes because of the FDL buffer memory capacity is limited when providing bigger spatial cache to reduce high capacity dividing into groups rises.
Control the connected mode of cache module according to the offered load size based on the optical buffer device of load selection, can reduce the competition resolution ability of light Packet Switch Node buffer memory total demand and raising node, fully improve the utilance of fibre delay line, well solved the ports competition problem of optical packet switch network, limited FDL buffer efficiency is improved, thereby the competition that improves the OPS node solve performance.
Description of drawings
Fig. 1 configuration is based on the OPS node structure schematic diagram of load selective light buffer storage
Fig. 2 exports segmentation and shares formula output FDL cache-time control chart
Fig. 3 output/feedback is shared formula FDL buffer storage and is operated in connection layout when exporting mode of extension
Fig. 4 output/feedback is shared formula FDL buffer storage and is operated in connection layout when feeding back shared model
Fig. 5 arrives light and divides group selection cache mode flow chart
Embodiment
OPS server node solves module by input port, switching matrix, control module, output port and FDL competition and forms, input port will amplify and regenerate from the grouping of same fiber optic wavelength channel, the packets headers of grouping is separated with payload, extract packets headers information and give control module, carry out routing addressing and control scheduling etc., control module is responsible for handling packets headers information and all instructions of issue, and switching matrix is configured.Switching matrix is the core component of OPS core node, finishes the function of exchange of payload according to the instruction of control module.It is the grouping that same output port wavelength channel failure is gone in the buffer memory competition that the FDL competition solves module, makes to be grouped in follow-up scheduling time by time delay and has an opportunity to be scheduled on the idle output port.FDL competition of the present invention solves the shared formula FDL cache module of module employing output/feedback can be according to arriving load size regulating and controlling cache mode, business works in the output buffers pattern when hanging down load, bigger buffer memory capacity is provided and guarantees that packet delay is less, when business load is higher, buffer storage is competed resolution ability, the low packet loss ratio performance when guaranteeing high capacity with the buffer storage that can be recycled by the light buffer memory that the feedback buffer memory increases the OPS node.
Below in conjunction with accompanying drawing technical scheme of the present invention is described further.
Figure 1 shows that output/feedback that configuration is selected based on load shares the OPS node optical buffer storage structural representation of selecting based on load of formula FDL buffer storage, by the shared formula FDL cache module of output/feedback, optical multiplexer/demodulation multiplexer, optical switch, optical combiner, the shared formula of segmentation export the FDL cache module, switching control module is formed.OPS server node (being called for short the OPS node), the part output port connects optical multiplexer by switching control module, share formula FDL module (be called for short output/feedback and share formula FDL module) input by optical multiplexer and output/feedback of selecting based on load and link to each other, this output/feedback is shared formula FDL cache module a plurality of input/output end ports (as M).Each input/output end port provides different time delays respectively, each output port connects an optical fiber, every optical fiber is divided into several sections (as the B sections), every period unit interval postpones to be D, the optical fiber of every section upper level can be shared all optical fiber of its next stage, has realized that each output port configuration capacity is BD.Every optical fiber cable length disposes in the mode that increases progressively a chronomere shares whole M output port, the time delay granularity that is M input/output end port is respectively D, 2D, 3D ..., MD, output/feedback is shared formula FDL module output and is all disposed a three-dimensional optical switch by fibre delay line, by the switching of optical switch, output/feedback is shared formula FDL cache module and is operated in the output mode of extension, or the feedback shared model.The fibre delay line of FDL cache module output connects the input of optical switch respectively, the output 1 of optical switch connects optical demultiplexer, and close the road by N output port of optical combiner and OPS node, the buffer memory capacity of expansion OPS node output port, the output 2 of optical switch feeds back the input port (being the switching control module input) that connects the OPS node respectively, further increases OPS nodal cache ability with the light cache way that can be recycled.OPS node switching control module is according to arriving which output whether business load surpasses threshold value decision selective light switch.When the traffic load values that arrives the OPS node is lower than business load threshold value thresholding H, the output port that output/feedback is shared formula FDL cache module is communicated with optical demultiplexer by optical switch output 1, be connected with the optical combiner of the output port of OPS node, buffer storage is operated under the expansion output shared buffer memory pattern, as the dilatation module of OPS node output buffers, play the output buffers ability effect that increases node; When being increased to, business load business load threshold value thresholding H is set when (as according to the desirable medium load value of emulation H), OPS node control optical switch switches, the output 2 of optical switch is communicated with, output 1 disconnects, each output port feedback that output/feedback is shared formula FDL cache module is connected to OPS node input port, by feedback FDL output of OPS node and input port is communicated with.All the other output ports of switching control module (1,2 ... B) by fibre delay line by segmentation share the output of formula output FDL cache module buffer memory (1,2 ... N).
If i.e. ρ 〉=H, then M FDL buffer memory output port of the optical buffer device of selecting based on load is communicated with M of OPS node by the optical switch feedback and expands input port, allow the grouping circulation of conflict repeatedly to enter optical buffer device, the optical buffer device of selecting based on load works in feedback shared buffer memory pattern, the conflict resolution ability when helping increasing high capacity; If ρ＜H, then M FDL buffer memory output port of the optical buffer device of selecting based on load is connected the optical combiner of OPS node with optical demultiplexer by optical switch, the capacity of output buffers is shared in the segmentation of expansion OPS node, and the optical buffer device of selecting based on load works in expansion output buffers pattern.
In the present invention, in order further to reduce the demand of OPS server node to the FDL buffer memory capacity, the output FDL buffer memory of each output port configuration adopts segmentation to share the formula configuration structure, the output buffers that is each output port adopts segmentation to share configuration FDL, the switching control module output is selected an optical fiber, this optical fiber is divided into the B section, be all optical fiber that the optical fiber of every section upper level just can be shared its next stage, fully realize sharing of fibre delay line on the single port, Figure 2 shows that the time relationship schematic diagram of this configuration structure.Every section optical fiber provides 1 time delay granularity D, and 2 sections optical fiber provide 2 time delay granularity 2D ..., B section optical fiber provides the delay relation of B time delay granularity BD.Time of delay, scale was D
1, D
2..., D
B, in the OPS node structure, the 1st output port D of OPS node
1 1, D
2 1..., D
B 1Expression, the 2nd output port D of OPS node
1 2, D
2 2..., D
B 2Expression ..., N output port D
1 N, D
2 N..., D
B NExpression, available time-delay value is D, 2D ..., BD, like this, the fibre delay line that the shared formula B section time span of segmentation is D is equivalent to
The section time span is the buffer memory capacity that the fibre delay line of D provides, and has greatly reduced the length of demand fibre delay line.
Be illustrated in figure 3 as the shared formula FDL buffer storage of output/feedback and be operated in connection layout when exporting mode of extension.Switching control module connects optical multiplexer, each input of sharing formula FDL module by optical multiplexer and output/feedback of selecting based on load links to each other, the shared formula FDL module output port of output/feedback connects optical demultiplexer by optical fiber separately, the output of each output port of optical demultiplexer is closed the road by N output port of optical combiner and OPS node, the 1st output port D of OPS
1 1, D
2 1..., D
B 1Expression, N output port D
1 N, D
2 N..., D
B NExpression, available time-delay value is D, 2D ..., BD.The buffer memory degree of depth of each output port of OPS node expands to B+M from B, and its buffer memory capacity increases.
Be illustrated in figure 4 as the shared formula FDL buffer storage of output/feedback and be operated in connection layout when feeding back shared model.By the switching of optical switch, each output port feedback that output/feedback is shared formula FDL cache module be connected to OPS node input port (1,2 ... M), by feedback FDL with output of OPS node and input port connection.At this moment the light buffer memory of OPS node is a mixed cache structure, i.e. formula FDL buffer memory is shared in OPS node output port configuration segmentation, has feedback to share the work of formula FDL cache module between the output/input port of OPS node, and its ability that manages conflict is stronger.
The cache mode of buffer storage control and select to come control switch with the business load size of statistics, decision are communicated to output port and still feed back and be communicated to the OPS input port, and its control procedure as shown in Figure 5.Light grouping to a newly arrived scheduling conflict, the control module statistics network arrives the business load ρ size of buffer storage and judges whether to exceed business load threshold value thresholding H, if ρ＜H, think that then network traffic load is lighter, the buffer memory output port of the shared formula FDL buffer storage of output/feedback closes the road by the output port of optical switch and optical demultiplexer and OPS node, output buffers expansion capacity as the OPS node uses, as shown in Figure 3, mixer by light, the buffer memory degree of depth of each output port of OPS node expands to B+M from B, and its buffer memory capacity increases.If arrive business load ρ 〉=H, think that then network traffic load is heavier, the pressure that conflict solves is bigger, the optical switch that output/feedback is shared formula FDL cache module switches, the output 2 of optical switch is communicated with, the M bar output delay line of output/reaction type FDL buffer storage directly is communicated with M input port of OPS node expansion, and output/feedback FDL buffer storage is operated under the feedback shared buffer memory pattern.
Specifically comprise the steps.
(1) arrive input port as new light grouping, Switching Module extracts the information that the packets headers analysis obtains according to control module, and OPS node statistics arrives the business load ρ of light grouping, sets business load threshold value thresholding H according to load value.
(2) packet scheduling conflict, scheduling configuration output port and channel, if run into conflict, whether the light grouping is at first sought has available segmentation to share formula output FDL cache module, if have, then switching control module shares this light packet scheduling with buffer memory on the formula output FDL cache module to the available segment of corresponding output end mouth.
Buffer memory is all unavailable if formula output FDL is shared in segmentation, further business load ρ is judged, when ρ 〉=H, the switches light switch, the output of output/feedback shared buffer memory module is connected with M expansion input port of OPS node, and the optical switch of optical buffer device is communicated with OPS expansion input.The feedback shared buffer memory can be used, the buffer scheduling grouping.Step (3) below changeing;
When ρ＜H, the switches light switch, the optical switch of optical buffer device is communicated with optical demultiplexer, the expansion output buffers, optical combiner by optical demultiplexer and OPS node output port, close the road with the shared reaction type FDL of the output port of OPS node, become expansion output and share the formula output buffers, change step (4).
(3) if output/feedback is shared idle FDL is arranged in the formula FDL module, contention packet is cached on the shortest relatively delay line of FDL length.If there is not available FDL, the light grouping is abandoned, change step (5).
(4) if segmentation is shared in the formula output FDL buffer memory idle FDL is arranged, the light packet scheduling of competing is cached to the shortest relatively available segment of FDL length shares in the formula output FDL cache module.If there is not available FDL, the light grouping is abandoned, change step (5).
(5) the buffer scheduling switching controls flow process that output/feedback is shared formula FDL device is withdrawed from switching control module control, prepares to handle the next grouping that arrives.
By with the contrast of general output type FDL buffer memory, in same buffer memory depth B (being the hop count of optical fiber) Under the condition, the segmented of wanting required for the present invention is shared output FDL buffer memory number and is obviously reduced. It is dark for example to work as buffer memory During degree B=8, degeneracy output type FDL buffer memory and the shared output buffer memory of segmentation fiber unit length are respectively 36 and 8, And along with the increase of B, this advantage that FDL demand total length is diminished further enlarges. The present invention proposes A kind of be applied to light packet-switched core network node select output/feedback sharing fiber delay line cache based on load Constructional device. This device changes FDL buffer memory mode of operation flexibly according to the business load size, low professional negative Be downloaded to the effect of exporting buffer memory capacity when reaching as expansion, when high business load, make as the feedback shared buffer memory With, feedback system allows the light grouping repeatedly to enter the buffer memory that the feedback fibre delay line solves the competition failed packet, has Effect ground has reduced the packet loss of grouping. Simultaneously, output FDL buffer memory is shared in the basic segmentation of output port configuration, Therefore its capacity can reduce the average delay of grouping much smaller than the demand of traditional output type FDL buffer memory, Improve the FDL utilization rate, reduce the switching node cost.
Claims (5)
1, a kind of OPS node optical buffer storage of selecting based on load, it is characterized in that, this device comprises, optical multiplexer/demodulation multiplexer, optical switch, optical branching device/mixer, share formula FDL module based on output/feedback that load is selected, formula output FDL cache module and switching control module are shared in segmentation, the part output of OPS node connects optical multiplexer by switching control module, sharing formula FDL module input by optical multiplexer and output/feedback links to each other, output/feedback is shared formula FDL module output and is connected the optical switch input by fibre delay line, the output 1 of optical switch connects optical demultiplexer, and close the road by the remainder output port of optical combiner and OPS node, connect segmentation and share formula output FDL cache module; The output 2 of optical switch feeds back the input port that connects the OPS node respectively, and whether switching control module surpasses threshold value thresholding decision selective light output switching terminal 1 or 2 connections according to arriving traffic load values.
2, OPS node optical buffer storage according to claim 1 is characterized in that, when arriving traffic load values and be lower than the threshold value thresholding, optical switch output 1 is communicated with optical demultiplexer, is connected with the optical combiner of the output port of OPS node; When arriving traffic load values more than or equal to the threshold value thresholding, OPS node control optical switch switches, and the output 2 of optical switch is communicated with, and the output port feedback that output/feedback is shared formula FDL cache module is connected to OPS node input port.
3, OPS node optical buffer storage according to claim 1 is characterized in that, output/feedback is shared each output port of formula FDL module and connected an optical fiber, and with the every average segmentation of optical fiber, the optical fiber of every section upper level is shared all optical fiber of its next stage.
4, OPS node optical buffer storage according to claim 1, it is characterized in that, the part output port of switching control module is shared formula FDL cache module of segmentation and the optical combiner identical with the buffer memory depth number of B by the fibre delay line allocating cache degree of depth, the remainder output port of switching control module is shared formula FDL module by optical multiplexer and output/feedback and is linked to each other, output/feedback is shared formula FDL module M input/output end port, D is provided respectively, 2D, 3D, the delay granularity of MD provides expansion light buffer memory by optical multiplexer for N output port of OPS node.
5, a kind of OPS node optical caching method of selecting based on load is characterized in that, comprises the steps:
(1) OPS node statistics arrives the business load ρ of light grouping, and sets business load threshold value thresholding H;
Whether (2) the light grouping is sought has available segmentation to share formula output FDL cache module, if having, then switching control module shares this light packet scheduling with formula output FDL cache module to available segment; If no, when ρ 〉=H, the switches light switch, the output of output/feedback being shared the formula cache module is connected with the expansion input port of OPS node, and commentaries on classics step (3) is when ρ＜H, the switches light switch is connected by the optical combiner of optical demultiplexer with OPS node output port, changes step (4);
(3) if output/feedback is shared idle FDL is arranged in the formula FDL module, the light grouping cache to the shortest delay line of FDL length, if there is not idle FDL, is abandoned the light grouping;
(4) if segmentation is shared in the formula output FDL buffer memory idle FDL is arranged, the light packet scheduling is shared in the formula output FDL cache module to the shortest available segment of FDL length,, the light grouping is abandoned if there is not available FDL.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200910190827A CN101656900B (en) | 2009-09-11 | 2009-09-11 | OPS node optical buffering device and method based on load selection |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200910190827A CN101656900B (en) | 2009-09-11 | 2009-09-11 | OPS node optical buffering device and method based on load selection |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101656900A true CN101656900A (en) | 2010-02-24 |
| CN101656900B CN101656900B (en) | 2012-09-26 |
Family
ID=41710942
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN200910190827A Active CN101656900B (en) | 2009-09-11 | 2009-09-11 | OPS node optical buffering device and method based on load selection |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101656900B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104219591A (en) * | 2013-05-30 | 2014-12-17 | 华为技术有限公司 | Data packet switching method and device |
| CN113556296A (en) * | 2021-05-27 | 2021-10-26 | 阿里巴巴新加坡控股有限公司 | Scheduling method, scheduling device, electronic equipment and storage medium |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1717113A (en) * | 2004-07-02 | 2006-01-04 | 电子科技大学 | Photoelectric mixed exchanging struture supporting light burst/group exchanging |
| CN1710839B (en) * | 2005-06-08 | 2010-04-28 | 北京交通大学 | Method for improving optical-fiber type complete optical buffer register characteristic |
| CN1866809B (en) * | 2006-06-22 | 2010-04-21 | 上海交通大学 | Optical buffer with adjustable delay time and output wavelength |
-
2009
- 2009-09-11 CN CN200910190827A patent/CN101656900B/en active Active
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104219591A (en) * | 2013-05-30 | 2014-12-17 | 华为技术有限公司 | Data packet switching method and device |
| CN104219591B (en) * | 2013-05-30 | 2017-12-01 | 华为技术有限公司 | A kind of packet-switching method and device |
| CN113556296A (en) * | 2021-05-27 | 2021-10-26 | 阿里巴巴新加坡控股有限公司 | Scheduling method, scheduling device, electronic equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101656900B (en) | 2012-09-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Danielsen et al. | Optical packet switched network layer without optical buffers | |
| TWI552536B (en) | Optical data center network system and optical switch | |
| EP2962417B1 (en) | System and method for steering packet streams | |
| CN101778047B (en) | Method for resolving optical switching grouping competition and photoelectric hybrid cache device | |
| CN102843294A (en) | Buffer device and buffer method for feeding back shared light based on FDL (fiber delay line) loop | |
| CN100525163C (en) | Wave-division multiplex optical exchanging system | |
| CN100514933C (en) | Communicating network exchanging system and controlling method based on parallel buffer structure | |
| CN101656900B (en) | OPS node optical buffering device and method based on load selection | |
| CN101370315A (en) | Sharing type optical packet switching node structure based on inside wavelength | |
| CN101621719B (en) | Switch processor matched with core node of hybrid optical switching network | |
| CN100421420C (en) | Method for dispatching variable length data packet queue in crossbar switching matrix | |
| CN100428660C (en) | Optical burst exchange node with internal acceleration | |
| CN100561980C (en) | Support the switching method and the node structure of burst packet and IP grouping | |
| CN110430486B (en) | Boundary switching node based on integrated hybrid optical switching network | |
| CN101193049B (en) | A core node switching structure for integrated edge node | |
| CN114979844B (en) | Optical switch and data communication system | |
| CN102427567A (en) | Asynchronous multi-wavelength mesh network adaptive node system based on optical packet switching | |
| CN101217335B (en) | A circumvention method of simple point reentrance of convertible wavelength intelligent optical network integrated routing | |
| CN102111692A (en) | Optical burst switching (OBS) channel scheduling method based on optical buffer (OB) | |
| Yan et al. | A novel scalable and low latency hybrid data center network architecture based on flow controlled fast optical switches | |
| CN1175597C (en) | Buffer-storage-free photo-exchanging matrix based on sharing wavelength convertor | |
| CN103986985B (en) | Based on network code and FDL loop feedback shared buffer memory devices and its control method | |
| CN1190915C (en) | Light-controlled light exchange structure adopting multiband mark | |
| CN104468414A (en) | Method for improving effective availability of links in optical burst switch network | |
| CN1909741A (en) | Wavelength/wave band sharing conversion structure |
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 |