CN101494611A - Method and apparatus for regulating load share - Google Patents
Method and apparatus for regulating load share Download PDFInfo
- Publication number
- CN101494611A CN101494611A CNA2009101055354A CN200910105535A CN101494611A CN 101494611 A CN101494611 A CN 101494611A CN A2009101055354 A CNA2009101055354 A CN A2009101055354A CN 200910105535 A CN200910105535 A CN 200910105535A CN 101494611 A CN101494611 A CN 101494611A
- Authority
- CN
- China
- Prior art keywords
- link
- cryptographic hash
- flow rate
- correspondence
- binding relationship
- 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 relates to the field of network communication, in particular to a load sharing adjusting method and a load sharing adjusting device. The method comprises the following steps: flow rate of each link is measured; predicted flow rate that is corresponding to each hash value is estimated according to a binding relation between links and hash values as well as the flow rate of each link, and the total number of hash values is more than the number of links; the binding relation between the links and the hash values is adjusted according to the predicted flow rate that is corresponding to each hash value, thus improving the load sharing uniformity. When the technical proposal of the embodiment of the invention is adopted, as the total number of hash values is more than the number of links, the load sharing uniformity of the flow-based routing forwarding technique can be improved by adjusting the binding relation between links and hash values according to the flow rate of each link.
Description
Technical field
The present invention relates to network communication field, relate in particular to a kind of load balancing method of adjustment and device.
Background technology
In the network equipment, to a destination address physics or the logical links of several identical expenses can be arranged, converting flow sends on these several links in turn.The purpose of configuration equal-cost route or trunking (Trunk) or tunnel (Tunnel) is to realize load balancing between multilink, improves link utilization, avoids link load inhomogeneous.
Trunk is the meaning of trunking, it is exactly setting by configuration software, thereby 2 or a plurality of physical port are combined the path that becomes a logic be increased in bandwidth between switch and the network node, the bandwidth that will belong to these several ports merges, and provides a high bandwidth that exclusively enjoys times over separate port to port.
Link routing technology comprises packet-by-packet the routing retransmission technique and by the stream routing retransmission technique.
With Trunk is example, and packet-by-packet the routing retransmission technique is to be unit with the packet, and each packet can be selected the different physical pathways among the Trunk, satisfies the harmony of load balancing to the full extent.Packet-by-packet there is the risk of packet out-ordering in the routing retransmission technique, general less use.
Select forward-path by the stream routing retransmission technique based on every stream, same stream is transmitted from same physical pathway.The following load balancing algorithm of general at present employing: according to source and destination Internet protocol (Internet Protocol, IP) address, source and destination medium access control (Media AccessControl, MAC) information such as address obtains a Hash (hash) value, carries out the selection of forward-path according to the binding relationship of cryptographic Hash and Trunk port.Because above Hash source (such as the IP address) has very big randomness in network, can produce cryptographic Hash at random, and then can be implemented in the load balancing between a plurality of Trunk ports.And same stream has information such as identical address, thereby cryptographic Hash is identical, according to the binding relationship of cryptographic Hash and Trunk port, transmits from same physical pathway.When using by the stream routing retransmission technique, the uniformity of load balancing and discharge model strong correlation, load sharing condition is poor.
Summary of the invention
The purpose of the embodiment of the invention provides a kind of load balancing method of adjustment and device, to improve the uniformity by stream routing retransmission technique load balancing.
The purpose of the embodiment of the invention is achieved through the following technical solutions:
A kind of load balancing method of adjustment comprises:
Measure each link flow;
Estimate the predicted flow rate that each cryptographic Hash is corresponding according to link with binding relationship and described each link flow between Hash (hash) value, the sum of described cryptographic Hash is greater than described number of links;
According to the predicted flow rate adjustment link of described each cryptographic Hash correspondence and the binding relationship between the cryptographic Hash, the uniformity of load balancing is improved.
A kind of load balancing adjusting device comprises:
Measurement module is used to measure each link flow;
Computing module is used for estimating the predicted flow rate that each cryptographic Hash is corresponding according to link with binding relationship and described each link flow between the cryptographic Hash that the sum of described cryptographic Hash is greater than described number of links;
Adjusting module is used for according to the predicted flow rate adjustment link of described each cryptographic Hash correspondence and the binding relationship between the cryptographic Hash uniformity of load balancing being improved.
The sum of cryptographic Hash adopts the technical scheme of the embodiment of the invention, because greater than number of links, can improve the uniformity by stream routing retransmission technique load balancing according to the binding relationship between each link flow adjustment link and the cryptographic Hash.
Description of drawings
Fig. 1 is a load balancing method of adjustment flow chart in the one embodiment of the invention;
Fig. 2 is a load balancing adjusting device block diagram in the one embodiment of the invention;
Fig. 3 is a load balancing method of adjustment flow chart in the another embodiment of the present invention;
Fig. 4 is a load balancing adjusting device block diagram in the another embodiment of the present invention.
Embodiment
Below in conjunction with Fig. 1 and Fig. 2 one embodiment of the invention is described:
Fig. 1 is a load balancing method of adjustment flow chart in the one embodiment of the invention.This method comprises:
102, measure each link flow.
For instance, the link here can be that Trunk port, Tunnel tunnel, equal-cost route or other need the link of load balancing.Can regularly measure each link flow, measuring intervals of TIME can be fixed, and also can be disposed by order line.For instance, can measure once each link flow in per 5 minutes.Also can measure each link flow according to the changes in flow rate situation, for example, set a changes in flow rate threshold value, when the changes in flow rate between twice measurement reduces measuring intervals of TIME during greater than this threshold value, when the changes in flow rate between twice measurement increases measuring intervals of TIME during less than this threshold value.Can also be in conjunction with timing measurement and the mode of measuring according to the changes in flow rate situation, for example set the fixation measuring time interval and a changes in flow rate threshold value, at fixing each link flow of Measuring Time point measurement, and, between two fixing Measuring Time points, increase one or many and measure when the changes in flow rate between twice measurement during greater than the changes in flow rate threshold value.
104, estimate the predicted flow rate that each cryptographic Hash is corresponding according to link with binding relationship and each link flow between the cryptographic Hash, the sum of described cryptographic Hash is greater than described number of links.Described cryptographic Hash is to obtain according to information such as the IP address of message and/or MAC Address in sending out by circulation, and its sum is sent out the algorithm decision by what adopt by circulation.Binding relationship between link and the cryptographic Hash is used to carry out the selection of forward-path.For example, the sum that cryptographic Hash can be set is 2: 1 or higher with the ratio of described number of links, and 4 links are for example arranged, and add up to 8 or more of cryptographic Hash then can be set.
For instance, the predicted flow rate of estimating each cryptographic Hash correspondence with the arithmetic average method according to sum and described each link flow of the cryptographic Hash of each link bundling.For example, 4 links are arranged, 16 cryptographic Hash, the flow of each link is respectively T
1, T
2, T
3, T
4, the cryptographic Hash of each link bundling has 3,3,4,6 respectively, and the predicted flow rate of then estimating 3 cryptographic Hash is T
1The predicted flow rate of/3,3 cryptographic Hash is T
2The predicted flow rate of/3,4 cryptographic Hash is T
3The predicted flow rate of/4,6 cryptographic Hash is T
4/ 6.
106, according to the predicted flow rate adjustment link of each cryptographic Hash correspondence and the binding relationship between the cryptographic Hash, the uniformity of load balancing is improved.
For improving the uniformity of the load balancing of transmitting by stream routing, need make the dispersion degree minimum of the predicted flow rate of adjusting each link correspondence of back, therefore, need provide the inhomogeneity quantitative criteria of load balancing.The predicted flow rate of described each link correspondence is meant the predicted flow rate sum of all cryptographic Hash correspondences of each link bundling.
For instance, as standard, the more little load balancing of extreme difference is even more with the extreme difference of the predicted flow rate of each link correspondence.Described extreme difference is meant the maximum of predicted flow rate of each link correspondence and the difference between the minimum value.Equally, the variance of the predicted flow rate of each link correspondence or standard deviation also can be used as standard, and the more little load balancing of variance or standard deviation is even more, described variance
Wherein n is the sum of link, T
iBe the predicted flow rate of each link correspondence, T is the mean value of the predicted flow rate of each link correspondence.Described standard deviation is the arithmetic square root of described variance.
Those of ordinary skills can know that other amounts that can represent data discrete degree or data fluctuations degree can be the quantitative criteria of the uniformity coefficient of load balancing as the effect of adjusting the binding relationship between link and the cryptographic Hash.
Can adopt algorithms of different to make the dispersion degree minimum of the predicted flow rate of each link correspondence of adjustment back.
For example, possible binding relationship between exhaustive all links and the cryptographic Hash, calculate the dispersion degree of the predicted flow rate that each link is corresponding under the binding relationship possible between each described link and the cryptographic Hash respectively, according to wherein the adjustment link of dispersion degree minimum and the binding relationship between the cryptographic Hash.
Again for example, with the binding relationship between greedy algorithm adjustment link and the cryptographic Hash, make the dispersion degree minimum of the predicted flow rate of adjusting each link correspondence of back.For instance, calculate the mean value of all link flows, surpass maximum link and the maximum links of subaverage of mean value from measuring flow, by adjusting the cryptographic Hash of above-mentioned two link bundlings, to the predicted flow rate of two links of mean value adjustment, make the absolute value sum minimum of the deviation of these two links and mean value.Two links of flow and mean deviation maximum begin to adjust from adjusted each link again, repeat above-mentioned steps, up to can't further optimizing or the quantitative criteria of dispersion degree reaches the expectation threshold value of prior appointment.
Those of ordinary skills can know that other can handle the method for optimization problem, as simulated annealing etc., may be used to make the dispersion degree minimum of the predicted flow rate of adjusting each link correspondence of back.With the simulated annealing is example, and simulated annealing is a kind of general probabilistic algorithm, is used for looking in a big search space optimal solution of proposition, and simulated annealing can be decomposed into solution space, target function and initial solution three parts.In embodiments of the present invention, solution space is the ensemble space of the binding relationship between all links and the cryptographic Hash, and target function is the dispersion degree index, and initial solution is to adjust preceding link and the binding relationship between the cryptographic Hash.The iteration each time of simulated annealing can be decomposed into for four steps, and the first step is to produce function by one to produce a new explanation that is positioned at solution space from current separating, and new explanation is only to adjust the binding relationship of a cryptographic Hash to produce in the embodiment of the invention.Second step was that calculating is poor with the pairing target function of new explanation, i.e. new explanation is with respect to the variable quantity of the dispersion degree of last iteration.The 3rd step was to judge whether new explanation is accepted, the most frequently used acceptance criterion is the Metropolis criterion: if S ' item is accepted as the new current S of separating in Δ t '＜0, otherwise with probability exp (Δ t '/T) accept S ' as the new current S of separating, wherein Δ t is that target function is poor, T is a target function.The 4th step was to be determined when accepting when new explanation, with current the separating of new explanation replacement.At this moment, current separating realized iteration one time.Can begin the next round test on this basis.And be judged as when giving up when new explanation, then on former current basis of separating, continue the next round test.
Fig. 2 is a load balancing adjusting device block diagram in the one embodiment of the invention.This device comprises:
For instance, the method for on average sharing this link flow according to each cryptographic Hash of same link correspondence is estimated the predicted flow rate of each cryptographic Hash correspondence.For example, 4 links are arranged, 16 cryptographic Hash, the flow of each link is respectively T
1, T
2, T
3, T
4, the cryptographic Hash of each link bundling has 3,3,4,6 respectively, and the predicted flow rate of then estimating 3 cryptographic Hash is T
1The predicted flow rate of/3,3 cryptographic Hash is T
2The predicted flow rate of/3,4 cryptographic Hash is T
3The predicted flow rate of/4,6 cryptographic Hash is T
4/ 6.
Adjusting module 206 is used for according to the predicted flow rate adjustment link of described each cryptographic Hash correspondence and the binding relationship between the cryptographic Hash uniformity of load balancing being improved.
106 method is identical among the method for the binding relationship between described adjustment link and the cryptographic Hash and the described embodiment of Fig. 1.
Below in conjunction with Fig. 3 and Fig. 4 one embodiment of the invention is described:
Fig. 3 is a load balancing method of adjustment flow chart in the another embodiment of the present invention.This method comprises:
302, measure each link flow.
The method of measuring each link flow in the present embodiment is identical with the method for middle each link flow of measurement embodiment illustrated in fig. 1.
304,, estimate the predicted flow rate of each cryptographic Hash correspondence according to the binding relationship between link and the cryptographic Hash, described each link flow and each cryptographic Hash corresponding historical predicted flow rate.
For example, 4 links are arranged, 16 cryptographic Hash, this flow of each link is respectively T
01, T
02, T
03, T
04The binding relationship of each link and cryptographic Hash is T
01: H
1-H
3, T
02: H
4-H
6, T
03: H
7-H
10, T
04: H
11-H
16The predicted flow rate of each cryptographic Hash correspondence was HL when measured last time
1-HL
16The predicted flow rate of then estimating the H1 correspondence is T
01[HL
1/ (HL
1+ HL
2+ HL
3)], by that analogy, when calculating this measurement, the predicted flow rate of each cryptographic Hash correspondence.
Everybody can know ordinary skill, and other evaluation methods of quoting historical record also can be used to estimate the predicted flow rate of each cryptographic Hash correspondence.For example, the predicted flow rate of each cryptographic Hash correspondence of estimation when quoting over three measurements, the weighting ratio of predicted flow rate in each link flow of each cryptographic Hash of each link bundling estimated in the predicted flow rate weighting conversion of estimation when by certain weight past of each cryptographic Hash correspondence being measured for three times.Weight can be the fixed value that sets in advance, and also can be the variable according to each link flow, each link flow stability, measuring intervals of TIME or other parameter adjustments.Still with aforementioned 4 links, 16 cryptographic Hash are example, the predicted flow rate of each cryptographic Hash correspondence of estimation when quoting over twice measurement, and the weight of past twice measurement is respectively Q
1And Q
2The predicted flow rate of each cryptographic Hash correspondence was H when twice of past measured
1L
1-H
1L
16And H
2L
1-H
2L
16The predicted flow rate of then estimating the H1 correspondence is T
01[(Q
1H
1L
1+ Q
2H
2L
1)/(Q
1H
1L
1+ Q
1H
1L
2+ Q
1H
1L
3+ Q
2H
2L
1+ Q
2H
2L
2+ Q
2H
2L
3)], by that analogy, when calculating this measurement, the predicted flow rate of each cryptographic Hash correspondence.
306, according to the predicted flow rate adjustment link of each cryptographic Hash correspondence and the binding relationship between the cryptographic Hash, the uniformity of load balancing is improved.
The method of adjusting the binding relationship between link and the cryptographic Hash according to the predicted flow rate of each cryptographic Hash correspondence in the present embodiment with embodiment illustrated in fig. 1 in basic identical according to the method for corresponding predicted flow rate adjustment link of each cryptographic Hash and the binding relationship between the cryptographic Hash.
Fig. 4 is a load balancing adjusting device block diagram in the another embodiment of the present invention.This device comprises:
The method of the predicted flow rate of each cryptographic Hash correspondence of described estimation is identical with 304 methods among the described embodiment of Fig. 3.
Adjusting module 408 is used for according to the predicted flow rate adjustment link of described each cryptographic Hash correspondence and the binding relationship between the cryptographic Hash uniformity of load balancing being improved.
The method of the binding relationship between described adjustment link and the cryptographic Hash is identical with 306 methods among the described embodiment of Fig. 3.
The sum of cryptographic Hash adopts embodiment of the invention technical scheme, because greater than number of links, can improve the uniformity by stream routing retransmission technique load balancing according to the binding relationship between each link flow adjustment link and the cryptographic Hash.
One of ordinary skill in the art will appreciate that all or part of step that realizes in the foregoing description method is to instruct relevant hardware to finish by program, described program can be stored in the computer-readable recording medium, described storage medium can be ROM/RAM, disk or CD etc.
The above; only for the preferable embodiment of the present invention, but protection scope of the present invention is not limited thereto, and anyly is familiar with those skilled in the art in the technical scope that the present invention discloses; the variation that can expect easily or replacement all should be encompassed within protection scope of the present invention.Therefore, protection scope of the present invention should be as the criterion with the protection range of claim.
Claims (10)
1, a kind of load balancing method of adjustment is characterized in that, comprising:
Measure each link flow;
Estimate the predicted flow rate that each cryptographic Hash is corresponding according to link with binding relationship and described each link flow between the cryptographic Hash, the sum of described cryptographic Hash is greater than described number of links;
According to the predicted flow rate adjustment link of described each cryptographic Hash correspondence and the binding relationship between the cryptographic Hash, the uniformity of load balancing is improved.
2, method according to claim 1 is characterized in that, described link comprises:
Trunking Trunk port; Or
The tunnel; Or
Equal-cost route.
3, method according to claim 1 is characterized in that, described cryptographic Hash obtains according to the IP address and/or the MAC Address of message, and the binding relationship between link and the cryptographic Hash is used for the selection of forward-path.
4, method according to claim 1 is characterized in that, each link flow of described measurement comprises:
Regularly measure each link flow; And/or
Measure each link flow according to changes in flow rate.
5, according to the arbitrary described method of claim 1 to 4, it is characterized in that, describedly estimate the predicted flow rate that each cryptographic Hash is corresponding with binding relationship and described each link flow between the cryptographic Hash, comprising according to link:
The predicted flow rate of estimating each cryptographic Hash correspondence with the arithmetic average method according to sum and described each link flow of the cryptographic Hash of each link bundling;
According to the binding relationship between described link and the cryptographic Hash, described each link flow and each cryptographic Hash corresponding historical predicted flow rate, estimate the predicted flow rate of current each cryptographic Hash correspondence.
6, method according to claim 1 is characterized in that, described predicted flow rate according to described each cryptographic Hash correspondence is adjusted the binding relationship between link and the cryptographic Hash, and the uniformity of load balancing is improved, and comprising:
According to the predicted flow rate adjustment link of described each cryptographic Hash correspondence and the binding relationship between the cryptographic Hash, make the dispersion degree minimum of the predicted flow rate of adjusting each link correspondence of back.
7, method according to claim 6 is characterized in that, the described dispersion degree of adjusting the predicted flow rate of each link correspondence of back comprises following arbitrary:
Adjust the extreme difference of the predicted flow rate of each link correspondence of back;
Adjust the variance of the predicted flow rate of each link correspondence of back;
Adjust the standard deviation of the predicted flow rate of each link correspondence of back.
8, according to claim 6 or 7 described methods, it is characterized in that, described predicted flow rate according to described each cryptographic Hash correspondence is adjusted the binding relationship between link and the cryptographic Hash, makes the dispersion degree minimum of the predicted flow rate of adjusting each link correspondence of back, comprises following arbitrary:
Binding relationship between exhaustive all links and the cryptographic Hash calculates the described dispersion degree under the binding relationship between each described link and the cryptographic Hash respectively, according to wherein the adjustment link of dispersion degree minimum and the binding relationship between the cryptographic Hash;
With the binding relationship between greedy algorithm adjustment link and the cryptographic Hash, make the dispersion degree minimum of the predicted flow rate of adjusting each link correspondence of back;
With the binding relationship between simulated annealing adjustment link and the cryptographic Hash, make the dispersion degree minimum of the predicted flow rate of adjusting each link correspondence of back.
9, a kind of load balancing adjusting device is characterized in that, comprising:
Measurement module is used to measure each link flow;
Computing module is used for estimating the predicted flow rate that each cryptographic Hash is corresponding according to link with binding relationship and described each link flow between the cryptographic Hash that the sum of described cryptographic Hash is greater than described number of links;
Adjusting module is used for according to the predicted flow rate adjustment link of described each cryptographic Hash correspondence and the binding relationship between the cryptographic Hash uniformity of load balancing being improved.
10, device according to claim 7, it is characterized in that, also comprise logging modle, be used to write down each cryptographic Hash corresponding historical predicted flow rate, correspondingly, described computing module is used for estimating the predicted flow rate that each cryptographic Hash is corresponding according to link with binding relationship, described each link flow and described each cryptographic Hash corresponding historical predicted flow rate between the cryptographic Hash that the sum of described cryptographic Hash is greater than described number of links.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2009101055354A CN101494611B (en) | 2009-02-20 | 2009-02-20 | Method and apparatus for regulating load share |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2009101055354A CN101494611B (en) | 2009-02-20 | 2009-02-20 | Method and apparatus for regulating load share |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101494611A true CN101494611A (en) | 2009-07-29 |
| CN101494611B CN101494611B (en) | 2011-09-14 |
Family
ID=40925026
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2009101055354A Expired - Fee Related CN101494611B (en) | 2009-02-20 | 2009-02-20 | Method and apparatus for regulating load share |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101494611B (en) |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102025609A (en) * | 2009-09-18 | 2011-04-20 | 浙江省电力公司 | Packet message-based load balancing distribution system and method thereof |
| CN102821049A (en) * | 2012-08-15 | 2012-12-12 | 华为技术有限公司 | Method and device for forwarding message |
| CN103139094A (en) * | 2013-03-12 | 2013-06-05 | 华为技术有限公司 | Method and device for adjusting uneven load sharing |
| WO2014023023A1 (en) * | 2012-08-10 | 2014-02-13 | 华为技术有限公司 | Multi-cpu packet processing method and system, switching unit and board |
| CN104168220A (en) * | 2014-09-02 | 2014-11-26 | 福建星网锐捷网络有限公司 | Higig flow equalization processing method and device thereof |
| CN104283782A (en) * | 2014-10-14 | 2015-01-14 | 杭州华三通信技术有限公司 | Method and device for determining message forwarding paths in multi-protocol label switching network |
| CN106357547A (en) * | 2016-09-08 | 2017-01-25 | 重庆邮电大学 | Software-defined network congestion control algorithm based on stream segmentation |
| CN106506394A (en) * | 2016-11-21 | 2017-03-15 | 浙江宇视科技有限公司 | A kind of load-balancing method, system and device |
| CN108769111A (en) * | 2018-04-17 | 2018-11-06 | 平安科技（深圳）有限公司 | A kind of server connection method, computer readable storage medium and terminal device |
| CN109831388A (en) * | 2017-11-23 | 2019-05-31 | 中国电信股份有限公司 | Method and apparatus for optimizing flow load balance |
-
2009
- 2009-02-20 CN CN2009101055354A patent/CN101494611B/en not_active Expired - Fee Related
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102025609A (en) * | 2009-09-18 | 2011-04-20 | 浙江省电力公司 | Packet message-based load balancing distribution system and method thereof |
| WO2014023023A1 (en) * | 2012-08-10 | 2014-02-13 | 华为技术有限公司 | Multi-cpu packet processing method and system, switching unit and board |
| CN102821049A (en) * | 2012-08-15 | 2012-12-12 | 华为技术有限公司 | Method and device for forwarding message |
| CN103139094B (en) * | 2013-03-12 | 2017-04-12 | 华为技术有限公司 | Method and device for adjusting uneven load sharing |
| CN103139094A (en) * | 2013-03-12 | 2013-06-05 | 华为技术有限公司 | Method and device for adjusting uneven load sharing |
| CN104168220B (en) * | 2014-09-02 | 2017-05-17 | 福建星网锐捷网络有限公司 | Higig flow equalization processing method and device thereof |
| CN104168220A (en) * | 2014-09-02 | 2014-11-26 | 福建星网锐捷网络有限公司 | Higig flow equalization processing method and device thereof |
| CN104283782A (en) * | 2014-10-14 | 2015-01-14 | 杭州华三通信技术有限公司 | Method and device for determining message forwarding paths in multi-protocol label switching network |
| CN104283782B (en) * | 2014-10-14 | 2018-11-09 | 新华三技术有限公司 | The method and apparatus that message forwarding path is determined in multi-protocol label switching network |
| CN106357547A (en) * | 2016-09-08 | 2017-01-25 | 重庆邮电大学 | Software-defined network congestion control algorithm based on stream segmentation |
| CN106506394A (en) * | 2016-11-21 | 2017-03-15 | 浙江宇视科技有限公司 | A kind of load-balancing method, system and device |
| CN106506394B (en) * | 2016-11-21 | 2020-10-20 | 浙江宇视科技有限公司 | Load balancing method, system and device |
| CN109831388A (en) * | 2017-11-23 | 2019-05-31 | 中国电信股份有限公司 | Method and apparatus for optimizing flow load balance |
| CN108769111A (en) * | 2018-04-17 | 2018-11-06 | 平安科技（深圳）有限公司 | A kind of server connection method, computer readable storage medium and terminal device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101494611B (en) | 2011-09-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101494611B (en) | Method and apparatus for regulating load share | |
| Li et al. | QTCP: Adaptive congestion control with reinforcement learning | |
| CN107070794B (en) | Low-orbit information network optimal network benefit time delay constraint routing method | |
| CN102355670B (en) | Multichannel wireless mesh network channel distribution method | |
| CN102958057B (en) | A kind of dynamic spectrum optimization, Apparatus and system | |
| CN103428770B (en) | Flow allocation method in the multi-link parallel transmission of heterogeneous wireless network | |
| CN106993310B (en) | Integrated routing method and system based on Load-aware in a kind of wireless self-networking | |
| CN103051546B (en) | Delay scheduling-based network traffic conflict prevention method and delay scheduling-based network traffic conflict prevention system | |
| CN103780317A (en) | Double-threshold collaborative frequency spectrum perception method based on trust degree | |
| Gao et al. | An online energy allocation strategy for energy harvesting cognitive radio systems | |
| CN104486163A (en) | Single end available bandwidth measurement method and terminal adopting same | |
| US20070140148A1 (en) | Fast simulated annealing for traffic matrix estimation | |
| US8438300B2 (en) | Achieving distributed flow control via data conflation | |
| Ren et al. | A sdn-based dynamic traffic scheduling algorithm | |
| US8289851B2 (en) | Lightweight bandwidth-management scheme for elastic traffic | |
| Serati et al. | ADR-Lite: A Low-Complexity Adaptive Data Rate Scheme for the LoRa Network | |
| Huang et al. | Downlink power control for variable bit rate videos over multicell wireless networks | |
| Chiang et al. | Jointly optimal congestion control and power control in wireless multihop networks | |
| CN106105282A (en) | Link buffer zone state is utilized to carry out the system and method for traffic engineering | |
| Chan et al. | Adaptive duty cycling in sensor networks via Continuous Time Markov Chain modelling | |
| Rukmani et al. | Enhanced channel allocation scheme for cross layer management in wireless network based on interference management | |
| Zawodniok et al. | Energy-efficient rate adaptation MAC protocol for ad hoc wireless networks | |
| Rath et al. | A cross layer congestion control algorithm in wireless networks for TCP Reno-2 | |
| CN112019443A (en) | Multi-path data transmission method and device | |
| Chun et al. | A trusted NUMFabric algorithm for congestion price calculation at the Internet-of-Things datacenter |
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 |
Granted publication date: 20110914 Termination date: 20170220 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |