CN100488165C - Stream scheduling method - Google Patents
Stream scheduling method Download PDFInfo
- Publication number
- CN100488165C CN100488165C CNB2005100806926A CN200510080692A CN100488165C CN 100488165 C CN100488165 C CN 100488165C CN B2005100806926 A CNB2005100806926 A CN B2005100806926A CN 200510080692 A CN200510080692 A CN 200510080692A CN 100488165 C CN100488165 C CN 100488165C
- Authority
- CN
- China
- Prior art keywords
- stream
- pir
- cir
- ordered set
- token
- 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.)
- Active
Links
Images
Abstract
The method includes following steps: setting up assuring information rate (CIR) token barrel and peak information rate (PIR) token barrel for each flow; based on values of current CIR token and PIR token as well as whether there is data packet to be sent, the method adds flow to CIR ordered set, PIR ordered set, or inactive stream set in logic port belonged to; when initiating deployment for flows, the logic port dispatches flows in CIR ordered set and flows in PIR ordered set. The method raises efficiency for dispatching flows, and reduces wobble of operation.
Description
Technical field
(Asynchronous Transfer Mode, the ATM) flow control technique in the network relate in particular to the dispatching method of stream to the present invention relates to IP and asynchronous transmission.
Background technology
Dispatching technique mainly solves is to be that the unit carries out statistic multiplexing with the cell with the multichannel Business Stream how, makes it to become the multiplexing path of a circuit-switched data.On this data multiplex path, the cell that belongs to every road Business Stream can obtain quality assurance to a certain degree liberally, comprises shake, postpones bandwidth or the like.
Dispatching method of the prior art can be divided into following two kinds substantially:
One, based on the outbound port queue scheduling
Outbound port may corresponding several formations, and number of queues is not very big (about 8) generally.Its scheduling strategy that adopts is: in the port, first formation is that limit priority is other, and other formations are according to the weight allocation bandwidth.
Two, dispatch based on Business Stream
Corresponding situation is: flow amount very many (thousands of), and each stream has several formations (such as 4), and a plurality of streams will be from an outbound port output.Its scheduling strategy that adopts is: be to share the outbound port bandwidth according to the bandwidth ratio between the stream.
Above-mentioned first kind of dispatching method ratio be easier to realize, at the bottom of the cost, but its service quality that provides do not have in theory second kind good.Therefore, in order to satisfy the high request of user to service quality, second kind of stream scheduling method of main at present employing.Specifically comprise: each Business Stream of cycle polling, when serving active flow, bring and the guarantee information speed of this stream (Committed information rate, the CIR) token of the some that is directly proportional of bandwidth, and being added on the last remaining token; If current token value is long more than or equal to current bag, then send packet; And it is long that token value is deducted bag, long less than current bag up to the token value of this stream, just is polled to next stream.
According to the description of above-mentioned stream scheduling process as can be known, stream scheduling method of the prior art has following shortcoming:
(1) the stream dispatching efficiency is low; If adjacent its token value of several active flows is all long less than the current bag of respective queue team head, scheduler must spend several cycles and go to handle them, does not have the packet of stream to be sent out in these several cycles, causes the cycle waste, reduces dispatching efficiency; In the shortest time of delay, do not send and come the stream that the formation back can send, dispatch delay is strengthened, influence dispatching efficiency equally;
(2), can adopt the token value that makes minimum bandwidth greater than the long alternative of maximum bag for addressing the above problem; But after adopting this processing method, the token value of big bandwidth is too big, and its burst is chronic, and the business of other stream can not be inserted in the centre, makes the burst characteristic of stream scheduling bad, has strengthened professional shake.
Summary of the invention
The invention provides a kind of stream scheduling method, in order to solve the low problem of stream dispatching efficiency that exists in the prior art.
The inventive method comprises: to each stream guarantee information speed CIR token bucket and peak information rate PIR token bucket are set, and according to the size of the size of the current C IR token value that flows in the token bucket, PIR token value and whether there is packet to send, under stream is joined in the CIR ordered set of logic port, PIR ordered set or the set of non-active flow; Described CIR ordered set and PIR ordered set are set up doubly linked list according to the sequencing that stream joins in the set respectively; When logic port is initiated the stream scheduling, carry out the following step:
A, judge whether CIR ordered set is nonvoid set, if, then send a packet of the current stream of doubly linked list in the CIR ordered set, go to step C; Otherwise, continue the following step;
B, judge whether PIR ordered set is nonvoid set, if, then send a packet of the current stream of PIR ordered set doubly linked list, go to step C; Otherwise, execution in step C;
C, finish this scheduling flow.
According to said method of the present invention, described steps A further comprises:
A1, with the next one of doubly linked list in CIR ordered set stream as current stream;
The CIR token value of A2, the current stream that will be scheduled and the bag that the PIR token value deducts described packet respectively are long;
A3, whether judge CIR token value and PIR token value, if the CIR token value greater than 0 and there is packet to send, then is retained in current stream in the CIR ordered set greater than 0; If the CIR token value smaller or equal to 0, the PIR token value is greater than 0 and have packet to send, and then current stream is deleted from CIR ordered set, join in the PIR ordered set; If CIR token value and PIR token value all smaller or equal to 0 or the free of data bag send, current stream is deleted from CIR ordered set, join in the non-active flow set.
According to said method of the present invention, described step B further comprises:
A packet of the current stream of doubly linked list in B1, the transmission PIR ordered set, and the next one of doubly linked list flows as current stream in will gathering;
The CIR token value of B2, the current stream that will be scheduled and the bag that the PIR token value deducts this packet respectively are long;
B3, whether judge the PIR token value, if the PIR token value greater than 0 and there is packet to send, then is retained in current stream in the PIR ordered set greater than 0; Otherwise, current stream is deleted from PIR ordered set, join in the non-active flow set.
The initial CIR token value and the initial p IR token value of stream are 0; CIR weighted value and the PIR weighted value CIR token and the PIR token that periodically increase proportional quantity according to stream.
Described periodicity increases the CIR token and the PIR token of proportional quantity, concrete grammar is: the stream information storage list is set, the current C IR token value of storage flow, PIR token value, CIR weighted value and PIR weighted value, system cycle reads described stream information storage list, CIR weighted value and PIR weighted value according to stream are CIR token and the PIR token that CIR token value, PIR token value increase proportional quantity, and refresh the current C IR token value and the PIR token value of stream.
The initial sets of described stream is the non-active flow set of affiliated logic port; According to current C IR token value that flows in the stream information storage list and PIR token value described stream is joined in the CIR ordered set or PIR ordered set of affiliated logic port.
If the current C IR token value of stream is greater than 0 and have packet to send, then described stream is joined in the CIR ordered set of affiliated logic port; If the current C IR token value of stream smaller or equal to 0, the PIR token value is greater than 0 and have packet to send, under then described stream being joined in the PIR ordered set of logic port; Otherwise, this stream is retained in the non-active flow set of affiliated logic port.
Also store the first service queue pointer and the second service queue pointer in the described stream information storage list, respectively in order to indicate each stream in CIR ordered set or PIR ordered set doubly linked list on stream and next stream; When logic port flows the stream of dispatching in increase or deletion CIR ordered set or the PIR ordered set, revise the first service queue pointer and the second service queue pointer of related streams in this stream information storage list.
When logic port flows scheduling, determine the next one stream of current stream in CIR ordered set or PIR ordered set according to the described first service queue pointer or the second service queue pointer.
According to said method of the present invention, also be provided with the logic port message store table, be used for the current flow identification information of the CIR ordered set of stored logic port and/or PIR ordered set and indicate CIR ordered set and whether PIR ordered set is empty set; When logic port is initiated the stream scheduling, inquire about this logic port message store table, whether be that empty set is carried out corresponding scheduling flow according to the CIR ordered set and the PIR ordered set of described logic port correspondence.
According to said method of the present invention, also be included as stream one or more average information speed MIR token buckets are set, satisfy the MIR token bucket rate greater than the CIR token bucket rate and less than the PIR token bucket rate; And increase one or more MIR ordered sets at logic port, according to the size of the size of the size of the current C IR token value that flows in the token bucket, MIR token value, PIR token value with whether there is packet to send, under stream is joined in the CIR ordered set of logic port, MIR ordered set, PIR ordered set or the set of non-active flow; When logic port is initiated the stream scheduling, dispatch stream, the stream in the MIR ordered set and the stream in the PIR ordered set in the CIR ordered set successively.
The present invention has following beneficial effect:
1, the present invention is as long as exist active flow in the CIR ordered set of logic port or the PIR ordered set, the business of dispatching out successfully just, and compared with prior art, the inventive method does not have the waste on the cycle, and dispatching efficiency is higher;
2, each Business Stream in the inventive method priority scheduling CIR ordered set has guaranteed the minimum bandwidth that flows; When CIR ordered set is empty set, dispatch the active flow in MIR ordered set and the PIR ordered set again, make full use of remaining bandwidth;
3, the present invention periodically refreshes CIR token, MIR token and the PIR token of all streams, and according to the token value after refreshing, make stream redirect in CIR ordered set, MIR ordered set, PIR ordered set and the set of non-active flow, realized the automatic conversion of stream mode;
4, the present invention is in the CIR of scheduling logic port ordered set, during stream in MIR ordered set or the PIR ordered set, it is a packet once dispatching out current stream, and then the next one that goes in the set flows, a plurality of active flows in being about to gather are interted dispatches its packet in turn, just send the packet of this stream when being polled to an active flow in this dispatching method and the prior art always, inequality up to its token value less than the dispatching method till wrapping length or not having the packet transmission, therefore, the present invention has overcome long shortcoming of scheduling burst time in the prior art simultaneously, and professional shake is improved.
Description of drawings
Fig. 1 is one of stream mode conversion schematic diagram of the present invention;
Fig. 2 is two of stream mode conversion schematic diagram of the present invention;
Fig. 3 is three of stream mode conversion schematic diagram of the present invention;
Fig. 4 refreshes and stream mode flow path switch figure for the token value of the present invention's stream;
Fig. 5 flows one of scheduling flow figure for the present invention;
Fig. 6 flows two of scheduling flow figure for the present invention.
Embodiment
The present invention is provided with the stream information storage list, and the relevant information of stream of scheduling is stored in this stream information storage list.Relative static information comprising stream: as the CIR weighted value of stream and PIR weight power and the logic port of giving flow distribution; Whether the multidate information that also comprises stream: have packet to send, flow at the current state of affiliated logic port, the CIR token value of stream, the PIR token value and the service queue pointer of stream as stream, the field contents of concrete storage is shown in following table one:
Table one:
| Professional pointer | Professional pointer | The bag sign is arranged | The C/P sign | The CIR token value | The PIR token value | Affiliated port | The CIR weighted value | The PIR weighted value |
| pNxt | pPre | PKT_VLD | C/P | TC_CIR | TC_PIR | LPID | W_CIR | W_PIR |
The state that one embodiment of the invention will flow at logic port is divided into three kinds, belongs to three set respectively.Be specially: CIR ordered set, PIR ordered set and the set of non-active flow.In the table one, the C/P sign is in order to the current affiliated set of indication stream; Have bag mark P KT_VLD to receive the complete packet that belongs to this stream for " 1 " expression system, this stream has packet to wait to transmit, and distributes if belong to the packet of this stream, then with its reset; Professional pointer pNxt indicates the next one stream of current stream in affiliated set; Professional pointer pPre indicate current stream in affiliated set on a stream; These two professional pointers on-the-fly modify with the redirect of stream in CIR ordered set and PIR ordered set.
The redirect schematic diagram of stream in three set as shown in Figure 1, Fi represents i stream among the figure.Wherein be arranged with several stream in the CIR ordered set, if current stream in the set of non-active flow, stream has bag to send and the TC_CIR of stream〉0, then its is added in CIR ordered set; If current stream is in non-active flow set, stream has the TC_PIR of bag transmission and stream〉0〉TC_CIR, then it is joined in the PIR ordered set; If current stream is in PIR ordered set, stream has the TC_CIR of bag transmission and stream〉0, then it is deleted from PIR ordered set, join in the CIR ordered set.
Except CIR and PIR ordered set, also can add one or more medium rates ordered sets (MIR) again, increased several medium rates ordered sets like this after, the business of each stream is being interleave more uniformly be scheduled away.As shown in Figure 2, for having increased the stream redirect schematic diagram of a MIR ordered set, its corresponding token bucket speed satisfies: CIR speed＜MIR speed＜PIR speed.Fig. 3 is for using the stream redirect schematic diagram of two MIR ordered sets.Whether the redirect rule of stream in each set is the same, also be according to its corresponding token value size and have bag to send, and jumps to another set from a set.
With logic port CIR is set below and PIR ordered set is example, the refresh flow of CIR and PIR token value and the concrete scheduling process of stream are described in detail.
The initial sets of supposing stream is set at non-active flow set, the initial TC_CIR of stream and TC_PIR are set at zero, system cycle reads described stream information storage list, CIR weighted value and PIR weighted value according to stream increase the CIR token and the PIR token of proportional quantity for it, join respectively in the CIR token bucket and PIR token bucket of stream, and refresh current C IR token value and PIR token value in the CIR token bucket of stream and the PIR token bucket.Concrete token value refresh and stream mode flow path switch figure as shown in Figure 4, comprise the steps:
Step S21, stream refresh cycle start the stream refresh flow when arriving;
Step S22, flow variables is set is i, and the initial value of i is 0, represents the 1st stream in the stream information storage list;
The recorded information of i stream in step S23, the reading flow message store table comprises the current C IR token value (TC_CIR) of stream, the current PIR token value (TC_PIR) of stream, the CIR weighted value (W_CIR) of stream and the PIR weighted value (W_PIR) of stream;
Step S24, increase pre-configured weighted value W_CIR and W_PIR respectively, that is:
TC_CIR
After refreshing=TC_CI
Before R refreshes+ W_CIR*RFRS_INTERVAL;
TC_PIR
After refreshing=TC_PIR
Before refreshing+ W_PIR*RFRS_INTERVAL;
In the formula, RFRS_INTERVAL is the interval duration of stream refresh cycle;
TC_CIR and TC_PIR after step S25, basis add up realize the redirect of stream in each set, are specially:
If TC_CIR〉0 and have bag to be sent, when flowing in PIR ordered set or the set of non-active flow, then will flow i deletion from PIR ordered set or the set of non-active flow, join in the CIR ordered set, rewrite the relative recording information (multidate information of stream) in the stream information storage list simultaneously;
If TC_CIR≤0 and TC_PIR〉0 there is bag to be sent, when flowing in CIR ordered set or the set of non-active flow, then will flow i deletion from CIR ordered set or the set of non-active flow, join in the PIR ordered set, rewrite relative recording information in the stream information storage list (multidate information of stream) simultaneously;
If TC_CIR≤0 and TC_PIR≤0 or do not have bag and wait to send out, when flowing in CIR ordered set or PIR ordered set, then will flow i from CIR ordered set or the deletion of PIR ordered set, join in the non-active flow set, rewrite relative recording information in the stream information storage list (multidate information of stream) simultaneously;
Step S26, judge in the stream information storage list whether all streams all refresh and finish, if not, continue step S27, otherwise, execution in step S28;
Step S27, be transferred to the next record in the stream information storage list, repeat above-mentioned steps;
Step S28, finish this time refresh flow.
In the above-mentioned stream refresh flow, need the stream in CIR ordered set and the PIR ordered set be added or delete, the present invention sets up doubly linked list according to the sequencing that stream joins in CIR ordered set or the PIR ordered set, and determines the relative front and back order of the flow queue in the set with this.
Logic port is transferred mobility for convenience, the present invention also is provided with the logic port message store table, and being used for storing the current flow identification information of the CIR ordered set of whole logic ports and/or PIR ordered set or indication CIR ordered set and/or PIR ordered set is empty set.Shown in the concrete following table two of storage field:
| Whether the CIR set is empty sign | Whether the PIR set is empty sign | CIR set queue heads information | PIR set queue heads information |
When logic port is initiated the stream scheduling, inquire about this logic port message store table, carry out as Fig. 5 and the described scheduling flow of Fig. 6, specifically comprise the steps:
Step S31, logic port are initiated the stream scheduling;
Step S32, be address lookup logic port message store table, obtain recorded information with the logical port number;
Step S33, judge that whether CIR ordered set is empty set, if empty set is then carried out flow process shown in Figure 6, carries out the scheduling of flowing in the PIR ordered set; Otherwise, carry out the following step;
A packet of current stream flows the next one in this set as current stream in step S34, the scheduling CIR ordered set;
The TC_CIR and the TC_PIR of step S35, the current stream that will be scheduled deduct long data packet respectively;
Step S36, according to current TC_CIR and the TC_PIR of stream, realize the redirect of stream in three set, be specially:
If TC_CIR〉0 and have bag to wait to send out, then this stream is retained in the CIR ordered set, rewrite the logic port message store table, execution in step S37;
If TC_CIR≤0 and TC_PIR〉0 there is bag to wait to send out, should flow from CIR ordered set and delete, join in the PIR ordered set rewriting stream information storage list and logic port message store table, execution in step S37;
If TC_CIR≤0 and TC_PIR≤0 or do not have bag and wait to send out will flows from CIR ordered set and be deleted, join in the non-active flow set rewriting stream information storage list and logic port message store table, execution in step S37;
Step S37, finish this scheduling flow.
Fig. 6 is that CIR ordered set starts stream scheduling in the PIR ordered set during for empty set, comprises the steps:
Step S41, judge whether PIR ordered set is empty set, if empty set represents that this logic port has not had active flow, execution in step S45; Otherwise, continue the following step;
A packet of current stream flows the next one in this set as current stream in step S42, the scheduling PIR ordered set;
The TC_CIR and the TC_PIR of step S43, the current stream that will be scheduled deduct long data packet respectively;
Step S44, according to current TC_CIR and the TC_PIR of stream, realize the redirect of stream in set, be specially:
If TC_PIR〉0 and have bag to be sent, this stream is retained in the PIR ordered set, rewrite the logic port message store table, execution in step S45;
If TC_PIR≤0 or not have bag to be sent will flow from PIR ordered set and be deleted, join in the set of non-active flow rewriting stream information storage list and logic port message store table, execution in step S45;
Step S45, finish this scheduling flow.
For the stream scheduling method that has increased MIR ordered set, its scheduling process and above-mentioned CIR and PIR scheduling are similar, do not repeat them here.
Method of the present invention has guaranteed the minimum bandwidth of stream by each Business Stream in the priority scheduling CIR ordered set; When CIR ordered set is empty set, dispatch the active flow in MIR and the PIR ordered set again, make full use of remaining bandwidth.Simultaneously, the present invention is when the stream of scheduling in each ordered set, it is a packet once dispatching out current stream, and then the next one that goes in the set flows, a plurality of active flows in being about to gather are interted dispatches its packet in turn, overcome long shortcoming of scheduling burst time in the prior art, made professional shake obtain fine improvement.
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 of the present invention 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 stream scheduling method, it is characterized in that, comprise: guarantee information speed CIR token bucket and peak information rate PIR token bucket are set to each stream, and according to the size of the size of the current C IR token value that flows in the token bucket, PIR token value and whether there is packet to send, under stream is joined in the CIR ordered set of logic port, PIR ordered set or the set of non-active flow; Described CIR ordered set and PIR ordered set are set up doubly linked list according to the sequencing that stream joins in the set respectively; When logic port is initiated the stream scheduling, carry out the following step:
A, judge whether CIR ordered set is nonvoid set, if, then send a packet of the current stream of doubly linked list in the CIR ordered set, go to step C; Otherwise, continue the following step;
B, judge whether PIR ordered set is nonvoid set, if, then send a packet of the current stream of PIR ordered set doubly linked list, go to step C; Otherwise, execution in step C;
C, finish this scheduling flow.
2, the method for claim 1 is characterized in that, described steps A further comprises:
A1, with the next one of doubly linked list in CIR ordered set stream as current stream;
The CIR token value of A2, the current stream that will be scheduled and the bag that the PIR token value deducts described packet respectively are long;
A3, whether judge CIR token value and PIR token value, if the CIR token value greater than 0 and there is packet to send, then is retained in current stream in the CIR ordered set greater than 0; If the CIR token value smaller or equal to 0, the PIR token value is greater than 0 and have packet to send, and then current stream is deleted from CIR ordered set, join in the PIR ordered set; If CIR token value and PIR token value all smaller or equal to 0 or the free of data bag send, current stream is deleted from CIR ordered set, join in the non-active flow set.
3, method as claimed in claim 1 or 2 is characterized in that, described step B further comprises:
B1, with the next one of doubly linked list in PIR ordered set stream as current stream;
The CIR token value of B2, the current stream that will be scheduled and the bag that the PIR token value deducts this packet respectively are long;
B3, whether judge the PIR token value, if the PIR token value greater than 0 and there is packet to send, then is retained in current stream in the PIR ordered set greater than 0; Otherwise, current stream is deleted from PIR ordered set, join in the non-active flow set.
4, the method for claim 1 is characterized in that, the initial CIR token value and the initial p IR token value of stream are 0; CIR weighted value and the PIR weighted value CIR token and the PIR token that periodically increase proportional quantity according to stream.
5, method as claimed in claim 4, it is characterized in that, described periodicity increases the CIR token and the PIR token of proportional quantity, concrete grammar is: the stream information storage list is set, the current C IR token value of storage flow, PIR token value, CIR weighted value and PIR weighted value, system cycle reads described stream information storage list, CIR weighted value and PIR weighted value according to stream are CIR token and the PIR token that CIR token value, PIR token value increase proportional quantity, and refresh the current C IR token value and the PIR token value of stream.
6, method as claimed in claim 5 is characterized in that, the initial sets of described stream is the non-active flow set of affiliated logic port; According to current C IR token value that flows in the stream information storage list and PIR token value described stream is joined in the CIR ordered set or PIR ordered set of affiliated logic port.
7, method as claimed in claim 6 is characterized in that, if the current C IR token value of stream is greater than 0 and there is packet to send, under then described stream being joined in the CIR ordered set of logic port; If the current C IR token value of stream smaller or equal to 0, the PIR token value is greater than 0 and have packet to send, under then described stream being joined in the PIR ordered set of logic port; Otherwise, this stream is retained in the non-active flow set of affiliated logic port.
8, method as claimed in claim 5, it is characterized in that, also store the first service queue pointer and the second service queue pointer in the described stream information storage list, respectively in order to indicate each stream in CIR ordered set or PIR ordered set doubly linked list on stream and next stream; When logic port flows the stream of dispatching in increase or deletion CIR ordered set or the PIR ordered set, revise the first service queue pointer and the second service queue pointer of related streams in this stream information storage list.
9, method as claimed in claim 8 is characterized in that, when logic port flows scheduling, determines the next one stream of current stream in CIR ordered set or PIR ordered set according to the described first service queue pointer or the second service queue pointer.
10, the method for claim 1, it is characterized in that, also be provided with the logic port message store table, be used for the current flow identification information of the CIR ordered set of stored logic port and/or PIR ordered set and indicate CIR ordered set and whether PIR ordered set is empty set; When logic port is initiated the stream scheduling, inquire about this logic port message store table, whether be that empty set is carried out corresponding scheduling flow according to the CIR ordered set and the PIR ordered set of described logic port correspondence.
11, the method for claim 1 is characterized in that, also is included as stream one or more average information speed MIR token buckets are set, and satisfies the MIR token bucket rate greater than the CIR token bucket rate and less than the PIR token bucket rate; And increase one or more MIR ordered sets at logic port, according to the size of the size of the size of the current C IR token value that flows in the token bucket, MIR token value, PIR token value with whether there is packet to send, under stream is joined in the CIR ordered set of logic port, MIR ordered set, PIR ordered set or the set of non-active flow; When logic port is initiated the stream scheduling, dispatch stream, the stream in the MIR ordered set and the stream in the PIR ordered set in the CIR ordered set successively.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2005100806926A CN100488165C (en) | 2005-07-06 | 2005-07-06 | Stream scheduling method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2005100806926A CN100488165C (en) | 2005-07-06 | 2005-07-06 | Stream scheduling method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1893389A CN1893389A (en) | 2007-01-10 |
| CN100488165C true CN100488165C (en) | 2009-05-13 |
Family
ID=37597911
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB2005100806926A Active CN100488165C (en) | 2005-07-06 | 2005-07-06 | Stream scheduling method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN100488165C (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101674482B (en) * | 2009-09-25 | 2011-05-11 | 上海大学 | Method for optimized dispatching of extension type video flow in partially observational Markovian decision process |
| CN102036131A (en) * | 2009-09-28 | 2011-04-27 | 中兴通讯股份有限公司 | End-to-end QoS (Quality of Service) control method and device based on GPON (Gigabit Passive Optical Network) system |
| CN102546423B (en) * | 2012-01-20 | 2014-10-29 | 福建星网锐捷网络有限公司 | Method and device for queue scheduling and network device |
| CN103533660B (en) * | 2012-07-02 | 2019-02-01 | 中兴通讯股份有限公司 | Downlink LCH dispatching method and device |
| CN117579564A (en) * | 2024-01-19 | 2024-02-20 | 成都智明达电子股份有限公司 | Multi-path flow scheduling system and method based on FPGA and token bucket algorithm |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5596576A (en) * | 1995-11-03 | 1997-01-21 | At&T | Systems and methods for sharing of resources |
| US20020087715A1 (en) * | 2000-12-28 | 2002-07-04 | Alcatel | Marking determining device and related method |
| US20030179774A1 (en) * | 2002-03-25 | 2003-09-25 | Erlang Technology, Inc. | Method and apparatus for WFQ scheduling using a plurality of scheduling queues to provide fairness, high scalability, and low computation complexity |
| WO2003088605A1 (en) * | 2002-04-12 | 2003-10-23 | Telecom Italia S.P.A. | Scheduling a shared resource among synchronous and asynchronous packet flows |
| CN1518296A (en) * | 2003-01-13 | 2004-08-04 | 华为技术有限公司 | Method of implementing integrated queue scheduling for supporting multi service |
| US6829649B1 (en) * | 2000-11-10 | 2004-12-07 | International Business Machines Corporation | Method an congestion control system to allocate bandwidth of a link to dataflows |
| WO2005006678A1 (en) * | 2003-06-27 | 2005-01-20 | Cisco Technology, Inc. | Methods and devices for flexible bandwidth allocation |
| WO2004086689A3 (en) * | 2003-03-21 | 2005-01-27 | Cisco Tech Ind | Method for prioritasing access by an access point and for the implementation of a simple 802.11e hcf (hybrid coordination function) |
-
2005
- 2005-07-06 CN CNB2005100806926A patent/CN100488165C/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5596576A (en) * | 1995-11-03 | 1997-01-21 | At&T | Systems and methods for sharing of resources |
| US6829649B1 (en) * | 2000-11-10 | 2004-12-07 | International Business Machines Corporation | Method an congestion control system to allocate bandwidth of a link to dataflows |
| US20020087715A1 (en) * | 2000-12-28 | 2002-07-04 | Alcatel | Marking determining device and related method |
| US20030179774A1 (en) * | 2002-03-25 | 2003-09-25 | Erlang Technology, Inc. | Method and apparatus for WFQ scheduling using a plurality of scheduling queues to provide fairness, high scalability, and low computation complexity |
| WO2003088605A1 (en) * | 2002-04-12 | 2003-10-23 | Telecom Italia S.P.A. | Scheduling a shared resource among synchronous and asynchronous packet flows |
| CN1518296A (en) * | 2003-01-13 | 2004-08-04 | 华为技术有限公司 | Method of implementing integrated queue scheduling for supporting multi service |
| WO2004086689A3 (en) * | 2003-03-21 | 2005-01-27 | Cisco Tech Ind | Method for prioritasing access by an access point and for the implementation of a simple 802.11e hcf (hybrid coordination function) |
| WO2005006678A1 (en) * | 2003-06-27 | 2005-01-20 | Cisco Technology, Inc. | Methods and devices for flexible bandwidth allocation |
Also Published As
| Publication number | Publication date |
|---|---|
| CN1893389A (en) | 2007-01-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102035732B (en) | Service scheduling method and device | |
| CN100488165C (en) | Stream scheduling method | |
| US7580348B2 (en) | Communication apparatus, control method, and computer readable information recording medium | |
| JP3252823B2 (en) | Time-based scheduler architecture and method for ATM networks | |
| US7684326B2 (en) | Methods and structures for improved buffer management and dynamic adaptation of flow control status in high-speed communication networks | |
| CN105264509B (en) | Adaptive interruption joint in converged network | |
| CN102387076B (en) | Shaping-combined hierarchical queue scheduling method | |
| JPH10276220A (en) | Method and device for scheduling cell transmission and scheduler | |
| JPH11340983A (en) | Scheduling circuit and method | |
| CN108984280B (en) | Method and device for managing off-chip memory and computer-readable storage medium | |
| CN102665284B (en) | Uplink service transmission scheduling method and terminal | |
| JP2009147833A (en) | Buffer device using token bucket, and program | |
| WO1993015571A1 (en) | Smoothing delay-sensitive traffic offered to atm networks | |
| CN100499580C (en) | Deficit polling method of low delay dithering small burst | |
| CN105656807A (en) | Network chip multi-channel data transmission method and transmission device | |
| CN103379038A (en) | Traffic scheduling device and method | |
| CN103685062A (en) | Cache management method and device | |
| CN102546423B (en) | Method and device for queue scheduling and network device | |
| US7907632B2 (en) | Data output apparatus, communication apparatus and switch apparatus | |
| AU2024201415A1 (en) | Device, method, and system that virtualize a network | |
| JP2628701B2 (en) | Switch for prioritized information packets | |
| CN102025694A (en) | DSP (Digital Signal Processor) array based device and method for sending Ethernet data | |
| CN102143053A (en) | Method, device and system for transmitting data | |
| US7385987B1 (en) | Scheduling system and method for multi-level class hierarchy | |
| CN103701721A (en) | Message transmission method and device |
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 |