CN1518296A - Method of implementing integrated queue scheduling for supporting multi service - Google Patents
Method of implementing integrated queue scheduling for supporting multi service Download PDFInfo
- Publication number
- CN1518296A CN1518296A CNA031003524A CN03100352A CN1518296A CN 1518296 A CN1518296 A CN 1518296A CN A031003524 A CNA031003524 A CN A031003524A CN 03100352 A CN03100352 A CN 03100352A CN 1518296 A CN1518296 A CN 1518296A
- Authority
- CN
- China
- Prior art keywords
- message
- queue
- service
- predefine
- sends
- 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
A method for implementing the integrated queue scheduling to support multiple services is disclosed. When predefined first priority message is added to queue, the speed at which the message in different class comes the queue relative to the first priority transmitted group is limited by the predefined first speed. If its speed is lower than or equal to the predefined first speed, it can come in the queue. If it is not it is rejected. When predefined first and second priority messages go out of the queue and the message in default queue is detected, their speed is limited by a predefined second speed. If their speed is lower than or equal to the predefined second speed, they can be transmitted. Or the predefined third message is transmitted.
Description
Technical field
The present invention relates to data communication field, relate in particular to the multiple services comprehensive array dispatching method of a kind of support.
Background technology
Under the complex environment of internet packet switching, network congestion is very common.The congested flow that makes can not in time obtain resource, is the source of causing service performance to descend, and congestedly might cause following negative effect: congested delay and the delay jitter that has increased message transmissions, too high delay can cause that message retransmits; The congested effective throughput of network that makes reduces, and causes the infringement of Internet resources; Congested aggravation can expend a large amount of Internet resources (particularly storage resources), irrational resource allocation even may cause system to be absorbed in the resource deadlock and collapse.Yet under packet switching and multi-user service and the complex environment deposited, congested is again common, and network takes place must manage and control it when congested, and common methods is to use queue technology, also can be described as queuing policy.Congestion management is exactly all will be classified from the message that an interface sends send into different formations, handles according to different strategies then; These strategies are used for handling the situation that requirement to bandwidth has exceeded the total bandwidth that network can provide.Congestion management is before congested generation and be failure to actuate, and its control enters the message flow of network when congested, to allow certain network traffics that the higher priority of other flows is relatively arranged.Be several frequently seen queuing policy below:
FIFO---fifo queue.Message is not classified, during speed that the speed that enters interface when message can send greater than interface, FIFO enters formation by the sequencing that message arrives interface by message, simultaneously, outlet message in formation team by the order of joining the team, advanced message will go out team earlier, laggard message with after go out team.
PQ---Priority Queues.Message is classified, message is sent into corresponding formation by the classification of message.The message that divides time-like to belong to higher priority queues will preferentially be sent, and the message of lower priority will be tried to be the first by the message of higher priority when generation is congested.
CQ---custom queuing.Message is classified, message is entered corresponding formation by the classification of message.When message goes out team, from formation, get a certain amount of message respectively in the bandwidth ratio of definition and send at interface.
WFQ---Weighted Fair Queuing.Message is classified by stream, and each stream is assigned to a formation.When going out group, the bandwidth that should occupy outlet by each stream of priority assignment of stream.The numerical value of priority is more little, and the gained bandwidth is few more.Otherwise the gained bandwidth is many more.
To sum up, the fifo queue technology is made no exception to miscellaneous service, together abandons during network congestion; Though the PQ queue technology has guaranteed the preferential of some business in a sense, owing to have only 4 formations, loaded service if preferential on the other hand business continues to send, may cause other business can't obtain scheduling forever very little on the one hand; The CQ queue technology has 16 formations, and each formation of round-robin scheduling can't guarantee the business of voice isopreference so again; WFQ compares preceding several queue technologies, improves to some extent, and number of queues increases relatively, and priority is high more, and weight is big more, and formation is weighted equity dispatching according to weight, but the same with the CQ queue technology, can't really guarantee the requirement of multiple business.Therefore, existing various queue technology be difficult to satisfy increasingly extensive multiple business and deposit, the requirement of Differentiated Services.
Summary of the invention
The purpose of this invention is to provide a kind of method that realizes supporting multiple services comprehensive queue scheduling, more definite the present invention is according to the characteristics of the miscellaneous service on the network, a kind of multi-service has been proposed and deposit, the method for the comprehensive queue scheduling of Differentiated Services.
The method according to this invention, the multiple services comprehensive queue scheduling implementation method of a kind of support, described method comprise that service message advances formation step and dequeue step, this service message advances the formation step and comprises:
11) when the first predefine prior message is joined the team, its inhomogeneous message is entered the first preferential transmission group respective queue and carry out the restriction of first set rate;
12) the second predefine prior message enters the second preferential set of queues that sends;
13) the 3rd predefined message enters default queue;
Wherein the first predefine prior message as described advance group speed be less than or equal to this first set rate then this first predefine prior message enter the first preferential transmission group respective queue that this message respective class is reserved, otherwise abandon;
Service message dequeue step comprises:
21) at first send the first preferential first predefine prior message that sends in the set of queues;
22) treat to send the second predefine prior message in the second preferential transmission set of queues again behind the first preferential transmission set of queues sky;
23) send the 3rd predefined message in the default queue at last;
24) detect default queue in real time whether described the 3rd predefined message is arranged;
In above-mentioned steps, if step 24) detect this default queue and have the 3rd predefined message then described first the or second predefine prior message to be gone out the speed that team sends when sending and carry out the restriction of second set rate when the described first predefine prior message or the second predefine prior message go out team, the first or second predefine prior message goes out group transmission rate and is less than or equal to described second set rate and then can goes out team's transmission as described;
First or the second predefine prior message goes out group speed and then sends the 3rd predefined message in the default queue greater than this second set rate as described;
If step 24) detect this default queue no third predefined message and then first the or second preferential message that sends is not gone out team and carries out the restriction of second set rate, repeated execution of steps 21)～22).
The concrete described first predefine prior message is EF (Expediated Forwarding) service message.
The concrete described second predefine prior message is AF (Assured Forwarding) service message.
Concrete described the 3rd predefined message is BE (Best-Effort Forwarding) service message.
Further described EF service message enters the first preferential set of queues that sends and at first sends, and the AF service message enters the second preferential set of queues that sends and treats that the first preferential set of queues sky that sends sends again, and the BE service message enters default queue and sends at last.
Wherein said first set rate restriction and the restriction of second set rate all adopt the token bucket rate algorithm to carry out rate limit, described first set rate is got the first token bucket Mean Speed, it is the bandwidth of described class configuration that this first token bucket Mean Speed adopts the user, the length of a maximum message segment is measured in burst, described second set rate is got the second token bucket Mean Speed, this second token bucket Mean Speed adopt each professional actual bandwidth of on interface, using and, maximum is no more than the maximum bandwidth reserved of this interface, and the length of a maximum message segment is measured in burst.
Comprise the step that determine to finish potential energy when in addition the AF service message being carried out queue operation, select when empty in the AF service queue the minimum message that finishes potential energy to go out team with convenient EF service queue and send, comprise following step:
Determine current potential energy of system;
Get this message before joining the team potential energy of system and the higher value of formation potential energy as the initial potential energy of described message;
According to the ratio of message length and institute's configured bandwidth again with initial potential energy and as the end potential energy of this message.
Concrete when the BE service message is advanced formation according to streams of feature differentiation such as the source IP address of message, purpose IP address, source port, destination interface, TOS (COS), agreement words, hash to each formation of BE, adopt the scheduling of quota polling algorithm to send during dequeue, wherein said quota is maximum bandwidth reserved remaining bandwidth of distributing after distributing to described EF service message and the required bandwidth of AF service message.
The present invention is by the mode of first priority transmission and token bucket rate restriction, both guaranteed the transmission quality of high priority message, the mode of the through-rate restriction problem having avoided the message of those Low Priority Queuings and can't obtain sending forever again simultaneously owing to there is not available bandwidth, reach Differentiated Services, guarantee multi-service and the problem of depositing simultaneously, bandwidth, delay, the delay jitter of message are guaranteed, reduce the network operation expense.
Description of drawings
Fig. 1 has described the flow chart of multi-business scheduling order of the present invention;
Fig. 2 has described the schematic diagram of the professional fair algorithm of AF of the present invention;
Fig. 3 has described the professional first priority scheduling of multiclass of the present invention and token bucket rate limits the flow chart that combines;
Embodiment
In order to allow those skilled in the art better understand method of the present invention, further describe the specific embodiment of the present invention below in conjunction with accompanying drawing.
In the present invention our predefine the message of three kinds of different COS, be respectively the first predefine prior message and the second predefine prior message and the 3rd predefined message, represent the service message of different business characteristics respectively, wherein, the described first predefine prior message is the message that need send at first, described second predefined message is lower than first predefine prior message transmission rank, the message of inferior transmission, described the 3rd predefined message then is the last message that sends, corresponding, described different predefined message classification enters different preferential transmission set of queues and sends, the first concrete predefine prior message enters the first preferential set of queues that sends and sends at first, the second predefine prior message enters the second preferential set of queues that sends and treats that the first preferential set of queues that sends sends for empty time, and the 3rd predefined message then enters default queue and sends at last.We know that the professional cardinal principle on the present network can be divided into: EF business, AF business and BE business.In the specific embodiment of the present invention, our the first predefine prior message of indication is the EF service message, the second predetermined prior message is the AF service message, the 3rd predefined message is the BE service message, below we are that example specifically describes the specific embodiment of the present invention with these several business just, business characteristic according to different business, adopt the method for the different business of different set of queues carryings among the present invention, the principle that while sends according to first priority when dequeue sends, guarantee that the high message of priority preferentially sends, in addition, the method that stresses a kind of queue scheduling in the present invention, message classification is entered team can adopt CBWFQ technology such as (class-based Weighted Fair Queuing) to carry out, no longer carefully state at this, only the array dispatching method that proposes with regard to the present invention describes in detail.
Under request in person referring to shown in Figure 1, characteristics according to each service message, the sending order of service message is that the EF service message sends at first in the present embodiment, the AF service message sends, the BE service message sends at last, promptly only behind EF service queue sky, could send the message of AF service queue, just send the BE service message behind the AF service queue sky, further set forth method of the present invention, we see the business characteristic of EF service queue first, and the EF service queue mainly comprises two-layer protocol message, speech business message etc. at real-time application.The EF service message requires to have strict time-delay, shake security mechanism and clock synchronization mechanism.And the main cause that causes packetization delay and packet jitter is a queueing delay, and queueing delay causes because of formation is very long, this situation can occur at network congestion point usually.When the grouping maximum arrival rate less than leaving speed, then can eliminate queueing delay basically.Therefore, the key issue of this service queue is as early as possible message to be sent out, and reduces delay and jitter to greatest extent.
Based on above analysis, because the two-layer protocol message limited amount, and undertaking the important task of maintenance link state, so do not doing any restriction when adopting two-layer protocol message to join the team among the present invention, the preferential principle that sends when going out group; For voice message, it is also very responsive to time delay and shake, can adopt token bucket rate limit when joining the team, the principle that first priority sends when going out group.
See the AF business characteristic below again, the AF business is at the business that requires to guarantee bandwidth when congested take place.Therefore, key issue is to guarantee bandwidth.
Because formation sends by bag, simulating GPS (broad sense processors sharing) system that therefore can't be desirable.The principle of GPS is that each is flowed in the logic query that all is placed on separately, once transmit infinitesimal data volume then for each non-empty queue, each is taken turns and all only transmits infinitesimal data volume, therefore in any limited time interval, all non-empty queue all can be accessed, whenever all is fair therefore.
We seek model approaching with GPS and that can realize to the AF service queue in the present embodiment.Concrete we introduce the notion of potential energy and come the state of tracking system, and the related formation potential energy of each connection (formation or class) comes service time that this formation obtains under the record interface congestion situation and sum service time of still needing; Potential energy of system of each interface conjunctionn writes down the time that this interface has been served since congested; Each wraps related initial potential energy and an end potential energy, is used for writing down the time of bag arrival and the time that the bag expectation is sent respectively.These potential energy all are the broad sense increasing function of a congestion period about the time.One when wrapping into group, pass through timestamp, long and the current potential energy of system of bag just can be unique determine a value, the message that the sending order of message in the unique decision current queue of this value, queue scheduling are chosen a minimum end potential energy during to the AF service queue sends.
Please referring to shown in the schematic diagram of the professional fair algorithm of Fig. 2 AF of the present invention, each formation has the formation potential energy (Pi) of oneself, each message has the beginning potential energy (SP) of oneself and finishes potential energy (TS), each interface has the potential energy of system (Ps) of oneself, we define the bandwidth that bi need guarantee for formation i simultaneously, b is the available bandwidth of this interface, and v is the formation number of non-NULL.The initial value of Ps and Pi all is 0, and the unit of potential energy is nanosecond in the present invention.
When an interface takes place when congested, all messages all need when sending through queue scheduling, briefly, need go into the operation of formation operation and dequeue.K the message that arrives with i formation is example, this moment, the SP of this message was this message join the team preceding Ps and the higher value of Pi, the ratio that the TS that calculates this message equals this message length and institute's configured bandwidth again with SP's and, this message is joined the team, corresponding Pi also increases to new TS value, so just guaranteed that each message in the interface queue all has the TS value of oneself, and the sole criterion that this TS value is weighed when to be message go out group.When physical layer comes an interruption to be ready for sending a message, this moment is because interface congestion, from formation, select a message to send, if be dispatched to the AF formation, the AF formation generally is made up of some formations, AF selects a minimum TS value from each formation message sends, please referring again to shown in Figure 2, the scheduling order of transmission of this message of numeral that marks in the message, wherein the TS value of numeral 1 this message of expression is minimum, sends at first, the TS value of 2 these messages of expression is taken second place, second sends, certainly, owing to advance a message or message of scheduling transmission in the AF formation at every turn, the TS value of all messages in the AF formation all will recomputate, so order of transmission also meets fair, in addition, because message sends with the available bandwidth speed of system, therefore, need update system potential energy, could guarantee that like this TS value all is issued less than the message of Ps, the work of simulating GPS system to greatest extent.
The priority of AF service queue dequeue is only second to the message of EF service queue, does not have message just can be dispatched to the AF service queue up to the EF formation.
Because the bandwidth between professional each formation of AF is can be adaptive, when certain formation did not have message in a certain moment AF formation, other formations in the AF formation can take this bandwidth in proportion.
AF has considered the time that message is joined the team among the present invention when going out group, and the size of message has guaranteed that to greatest extent message goes out the fairness of team, has reduced time delay.
See the BE business at last again, the BE business belongs to the data service that transmits as possible, as business of networking, ftp business.Generally, the BE service queue is not at needing to guarantee to delay time and the business of bandwidth, and therefore, such professional key issue is to guarantee to send out to such an extent that go out, and guarantees the relative fairness between the BE service queue simultaneously.
The a plurality of formations in BF service queue mountain are formed among the present invention, according to streams of feature differentiation such as the source IP address of message, purpose IP address, source port, destination interface, TOS (COS), agreement words, hash to each formation of BE before advancing formation.Guarantee fairness by polling algorithm during dequeue as far as possible.
Because the BE business belongs to the business that transmits as possible, the BE service queue adopts the quota polling algorithm when going out group, when being dispatched to the BE service queue, circulation sends quota from each formation, quota deficiency person waits for scheduling next time, described quota is maximum bandwidth reserved remaining bandwidth of distributing after distributing to described EF service message and the required bandwidth of AF service message, the concrete interface bandwidth as maximum is 75% of a total bandwidth, that then ought distribute to EF service message and AF business accounts for 60%, then has 15% quota can distribute to the BE service message.
In the present embodiment, it is the minimum formation of priority that the BE service queue is compared with the AF service queue with the EF service queue, just is scheduled at last.
Above-mentioned in the queue scheduling process, the EF service queue is preferential, the AF service queue takes second place, and the BE service queue is last.Like this can strict guarantee the bandwidth and the delay of high-priority queue, also may cause that the BE service queue can't obtain scheduling forever under some situation simultaneously.Yet consider that from the feature of miscellaneous service the AF business is to time-delay and insensitive.The present invention has adopted the mode of first priority scheduling and token bucket rate restriction combination to address this problem.The EF business when joining the team and EF service queue and AF service queue all will be when going out group through the token bucket rate restriction.
In the present embodiment, entering team when the EF business enters first preferential when sending set of queues, directly its inhomogeneous message being entered the first preferential transmission group respective queue that respective class reserves carries out first set rate and limits, wherein said first set rate is got the first token bucket Mean Speed, it is the bandwidth of such configuration that this first token bucket Mean Speed can adopt the user, and the length of a maximum message segment is measured in burst.If be less than or equal to this first token bucket rate, then enter the first preferential transmission group respective queue that respective class is reserved, otherwise abandon.
Please limit the flow chart that combines referring to professional first priority scheduling of Fig. 3 multiclass of the present invention and token bucket rate again, concrete steps were as follows when message went out team's transmission:
Step 1, judge the EF service queue is whether the described first preferential set of queues that sends has message, if any described EF service message, then this EF service message is gone out team's transmission and carry out the restriction of second set rate, wherein said second set rate is got another token bucket rate, here be designated as second token bucket rate in order to distinguish, in the present embodiment, the second token bucket Mean Speed adopt each the professional actual bandwidth on interface, use and, maximum is no more than maximum bandwidth reserved (because the BE service queue of interface, two-layer protocol formation message, physical layer overhead, ATM divides fixedly, and the shared expense of cell need take a part of bandwidth, the difference of interface actual bandwidth and this part bandwidth is maximum bandwidth reserved), the burst amount still is the length of a maximum message segment.EF business so as described goes out group transmission rate is less than or equal to described rate limit and then goes out team and send, otherwise execution in step 3;
Step 2, judge whether the AF service queue has message, be whether the described second preferential set of queues that sends has described AF service message, carry out the restriction of second set rate if any then its message being gone out team's transmission, concrete described second set rate still be the Mean Speed of second token bucket promptly adopt each professional actual bandwidth of on interface, using and, maximum is no more than the maximum bandwidth reserved of interface, be less than or equal to described second set rate and then continue to send when described AF service message goes out group transmission rate, otherwise execution in step 3;
Step 3, the BE service message in the scheduling default queue go out team and send;
Step 4, detect default queue in real time whether the BE service message is arranged, if have, then set by step 1 or the step 2 pair first predefine prior message or the second predefine prior message go out team and send and carry out the restriction of second set rate, otherwise do not carry out the restriction of second set rate, the principle that sends according to first priority sends described EF, AF service message.
Be understood that,, not only guaranteed the transmission quality of EF, the high message of AF isopreference level but also prevented the professional problem that can't obtain dispatching forever of BE owing to there is not available bandwidth through the restriction of first priority transmission of the present invention and token bucket rate.
The above only is a possible embodiments of the present invention, and is non-so promptly limit to interest field of the present invention, and the equivalence that all utilizations specification of the present invention and accompanying drawing content are done changes, and all reason is with being contained in the claim scope of the present invention.
Claims (10)
1, the multiple services comprehensive queue scheduling implementation method of a kind of support, described method comprises that service message advances formation step and dequeue step, it is characterized in that described service message advances the formation step and comprises:
11) when the first predefine prior message is joined the team, its inhomogeneous message is entered the first preferential transmission group respective queue and carry out the restriction of first set rate;
12) the second predefine prior message enters the second preferential set of queues that sends;
13) the 3rd predefined message enters default queue;
Wherein the first predefine prior message as described advance group speed be less than or equal to this first set rate then this first predefine prior message enter the first preferential transmission group respective queue that this message respective class is reserved, otherwise abandon;
Service message dequeue step comprises:
21) at first send the first preferential first predefine prior message that sends in the set of queues;
22) treat to send the second predefine prior message in the second preferential transmission set of queues again behind the first preferential transmission set of queues sky;
23) send the 3rd predefined message in the default queue at last;
24) detect default queue in real time whether described the 3rd predefined message is arranged.
2, the multiple services comprehensive queue scheduling implementation method of a kind of support according to claim 1, it is characterized in that also comprising: if described step 24) detect this default queue the 3rd predefined message is arranged, then the described first or second predefine prior message is gone out the speed that team sends when sending and carry out the restriction of second set rate when the described first predefine prior message or the second predefine prior message go out team, the first or second predefine prior message goes out then can go out team when group speed is less than or equal to described second set rate and send as described;
First or the second predefine prior message goes out group speed and then sends the 3rd predefined message in the default queue greater than this second set rate as described;
If step 24) detect this default queue no third predefined message and then first the or second preferential message that sends is not gone out team and carries out the restriction of second set rate, repeated execution of steps 21)～22).
3, the multiple services comprehensive queue scheduling implementation method of a kind of support according to claim 1 is characterized in that the described first predefine prior message is EF (ExpediatedForwarding) service message.
4, the multiple services comprehensive queue scheduling implementation method of a kind of support according to claim 1 is characterized in that the described second predefine prior message is AF (Assured Forwarding) service message.
5, the multiple services comprehensive queue scheduling implementation method of a kind of support according to claim 1 is characterized in that described the 3rd predefined message is BE (Best-Effort Forwarding) service message.
6, according to claim 3, the multiple services comprehensive queue scheduling implementation method of 4 or 5 described a kind of supports, it is characterized in that described EF service message enters the first preferential set of queues that sends and at first sends, the AF service message enters the second preferential set of queues that sends and treats that the first preferential set of queues sky that sends sends again, and the BE service message enters default queue and sends at last.
7, the multiple services comprehensive queue scheduling implementation method of a kind of support according to claim 1 is characterized in that described first set rate limits and the restriction of second set rate all adopts token bucket algorithm to carry out.
8, the multiple services comprehensive queue scheduling implementation method of a kind of support according to claim 7, it is characterized in that described first set rate gets the first token bucket Mean Speed, it is the bandwidth of described class configuration that this first token bucket Mean Speed adopts the user, and the length of a maximum message segment is measured in burst; Described second set rate is got the second token bucket Mean Speed, this second token bucket Mean Speed adopt each professional actual bandwidth of on interface, using and, maximum is no more than the maximum bandwidth reserved of this interface, the maximum length of a message is measured in burst.
9, the multiple services comprehensive queue scheduling implementation method of a kind of support according to claim 6, comprise the step of determining to finish potential energy when it is characterized in that the AF service message carried out queue operation, select when empty in the AF service queue the minimum message that finishes potential energy to go out team with convenient EF service queue and send, comprise following step:
Determine current potential energy of system;
Get this message before joining the team potential energy of system and the higher value of formation potential energy as the initial potential energy of described message;
According to the ratio of message length and institute's configured bandwidth again with initial potential energy and as the end potential energy of this message.
10, the multiple services comprehensive queue scheduling implementation method of a kind of support according to claim 6, when it is characterized in that the BE service message advanced formation according to source IP address, purpose IP address, source port, destination interface, TOS (COS), stream of agreement word feature differentiation of message, hash to each formation of BE, adopt the scheduling of quota polling algorithm to send during dequeue, wherein said quota is maximum bandwidth reserved remaining bandwidth of distributing after distributing to described EF service message and the required bandwidth of AF service message.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB200510028061XA CN100466593C (en) | 2003-01-13 | 2003-01-13 | Method of implementing integrated queue scheduling for supporting multi service |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB200510028061XA CN100466593C (en) | 2003-01-13 | 2003-01-13 | Method of implementing integrated queue scheduling for supporting multi service |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1518296A true CN1518296A (en) | 2004-08-04 |
| CN100466593C CN100466593C (en) | 2009-03-04 |
Family
ID=34281138
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB200510028061XA Expired - Fee Related CN100466593C (en) | 2003-01-13 | 2003-01-13 | Method of implementing integrated queue scheduling for supporting multi service |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN100466593C (en) |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN100370746C (en) * | 2004-10-10 | 2008-02-20 | 华为技术有限公司 | Scheduling method of mobile data service |
| CN100421428C (en) * | 2004-10-28 | 2008-09-24 | 华为技术有限公司 | Method for scheduling forward direction public control channel message |
| CN100488165C (en) * | 2005-07-06 | 2009-05-13 | 华为技术有限公司 | Stream scheduling method |
| CN102118314A (en) * | 2011-02-28 | 2011-07-06 | 华为技术有限公司 | Flow management method and management device |
| CN101557340B (en) * | 2009-05-07 | 2011-09-21 | 中兴通讯股份有限公司 | Method for realizing multilevel queue scheduling in data network and device |
| CN101151846B (en) * | 2005-02-28 | 2011-10-19 | 思科技术公司 | Method and system for shaping traffic in a parallel queuing hierarchy |
| CN101674242B (en) * | 2009-10-13 | 2011-12-28 | 福建星网锐捷网络有限公司 | Service message sending control method and device |
| CN102368741A (en) * | 2011-12-05 | 2012-03-07 | 盛科网络（苏州）有限公司 | Method supporting hierarchical queue scheduling and flow shaping and apparatus thereof |
| CN101692648B (en) * | 2009-08-14 | 2012-05-23 | 中兴通讯股份有限公司 | Queue scheduling method and system |
| CN101795173B (en) * | 2010-01-21 | 2012-11-21 | 福建星网锐捷网络有限公司 | Method for transmitting downlink data in wireless network and wireless access equipment |
| CN103973593A (en) * | 2014-05-09 | 2014-08-06 | 中国电子科技集团公司第三十研究所 | IP voice dispatching method |
| CN105573829A (en) * | 2016-02-02 | 2016-05-11 | 沈文策 | Method for fast processing high-traffic-flow data in system |
| CN106789731A (en) * | 2016-11-17 | 2017-05-31 | 天津大学 | Array dispatching method based on energy Internet service importance degree |
| CN113747597A (en) * | 2021-08-30 | 2021-12-03 | 上海智能网联汽车技术中心有限公司 | Network data packet scheduling method and system based on mobile 5G network |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| GB2338372B (en) * | 1998-06-12 | 2003-08-27 | Ericsson Telefon Ab L M | Architecture for integrated services packet-switched networks |
| US6914882B2 (en) * | 2001-03-30 | 2005-07-05 | Nokia, Inc. | Method and apparatus for improved queuing |
-
2003
- 2003-01-13 CN CNB200510028061XA patent/CN100466593C/en not_active Expired - Fee Related
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN100370746C (en) * | 2004-10-10 | 2008-02-20 | 华为技术有限公司 | Scheduling method of mobile data service |
| CN100421428C (en) * | 2004-10-28 | 2008-09-24 | 华为技术有限公司 | Method for scheduling forward direction public control channel message |
| CN101151846B (en) * | 2005-02-28 | 2011-10-19 | 思科技术公司 | Method and system for shaping traffic in a parallel queuing hierarchy |
| CN100488165C (en) * | 2005-07-06 | 2009-05-13 | 华为技术有限公司 | Stream scheduling method |
| CN101557340B (en) * | 2009-05-07 | 2011-09-21 | 中兴通讯股份有限公司 | Method for realizing multilevel queue scheduling in data network and device |
| CN101692648B (en) * | 2009-08-14 | 2012-05-23 | 中兴通讯股份有限公司 | Queue scheduling method and system |
| CN101674242B (en) * | 2009-10-13 | 2011-12-28 | 福建星网锐捷网络有限公司 | Service message sending control method and device |
| CN101795173B (en) * | 2010-01-21 | 2012-11-21 | 福建星网锐捷网络有限公司 | Method for transmitting downlink data in wireless network and wireless access equipment |
| CN102118314A (en) * | 2011-02-28 | 2011-07-06 | 华为技术有限公司 | Flow management method and management device |
| CN102368741A (en) * | 2011-12-05 | 2012-03-07 | 盛科网络（苏州）有限公司 | Method supporting hierarchical queue scheduling and flow shaping and apparatus thereof |
| CN103973593A (en) * | 2014-05-09 | 2014-08-06 | 中国电子科技集团公司第三十研究所 | IP voice dispatching method |
| CN103973593B (en) * | 2014-05-09 | 2017-03-15 | 中国电子科技集团公司第三十研究所 | A kind of ip voice dispatching method |
| CN105573829A (en) * | 2016-02-02 | 2016-05-11 | 沈文策 | Method for fast processing high-traffic-flow data in system |
| CN106789731A (en) * | 2016-11-17 | 2017-05-31 | 天津大学 | Array dispatching method based on energy Internet service importance degree |
| CN106789731B (en) * | 2016-11-17 | 2020-03-06 | 天津大学 | Queue scheduling method based on energy Internet service importance |
| CN113747597A (en) * | 2021-08-30 | 2021-12-03 | 上海智能网联汽车技术中心有限公司 | Network data packet scheduling method and system based on mobile 5G network |
Also Published As
| Publication number | Publication date |
|---|---|
| CN100466593C (en) | 2009-03-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CA2575869C (en) | Hierarchal scheduler with multiple scheduling lanes | |
| Ramabhadran et al. | Stratified round robin: A low complexity packet scheduler with bandwidth fairness and bounded delay | |
| US7782776B2 (en) | Shared weighted fair queuing (WFQ) shaper | |
| CN100463451C (en) | Multidimensional queue dispatching and managing system for network data stream | |
| CN1541474A (en) | System and method of differentiated queuing in routing system | |
| JP3306705B2 (en) | Packet transfer control device and scheduling method thereof | |
| JP2000049853A (en) | Rate warrant method and device by buffer management | |
| RU2005131960A (en) | ACCESS PERMISSION MANAGEMENT AND RESOURCE ALLOCATION IN THE COMMUNICATION SYSTEM WITH SUPPORT OF APPLICATION STREAMS WITH AVAILABILITY OF SERVICE QUALITY REQUIREMENTS | |
| CN100466593C (en) | Method of implementing integrated queue scheduling for supporting multi service | |
| CN101286949A (en) | Wireless Mesh network MAC layer resource scheduling policy based on IEEE802.16d standard | |
| CN102811176B (en) | A kind of data flow control method and device | |
| US7894347B1 (en) | Method and apparatus for packet scheduling | |
| Ramabhadran et al. | The stratified round robin scheduler: design, analysis and implementation | |
| WO2013025703A1 (en) | A scalable packet scheduling policy for vast number of sessions | |
| CN117395167A (en) | Service level configuration method and device | |
| US9128755B2 (en) | Method and apparatus for scheduling resources in system architecture | |
| Yaghmaee et al. | A model for differentiated service support in wireless multimedia sensor networks | |
| CN102769566A (en) | Method and device for configuring multilevel scheduling system, and method and device for changing configuration of multilevel scheduling system | |
| Shi | Token-based congestion control: Achieving fair resource allocations in P2P networks | |
| Victoria et al. | Efficient bandwidth allocation for packet scheduling | |
| Zhou | Resource allocation in computer networks: Fundamental principles and practical strategies | |
| Lenzini et al. | Bandwidth and latency analysis of modified deficit round robin scheduling algorithms | |
| CN101431802B (en) | Data transmission method, network node and network system | |
| Kogan et al. | Large profits or fast gains: A dilemma in maximizing throughput with applications to network processors | |
| Yan et al. | A Novel Packet Queuing And Scheduling Algorithm And Its Link Sharing Performance For Home Router |
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 | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20090304 Termination date: 20160113 |