CN101110767A - Accelerating method for asymmetric and multi-concurrency network - Google Patents
Accelerating method for asymmetric and multi-concurrency network Download PDFInfo
- Publication number
- CN101110767A CN101110767A CNA2007101111482A CN200710111148A CN101110767A CN 101110767 A CN101110767 A CN 101110767A CN A2007101111482 A CNA2007101111482 A CN A2007101111482A CN 200710111148 A CN200710111148 A CN 200710111148A CN 101110767 A CN101110767 A CN 101110767A
- Authority
- CN
- China
- Prior art keywords
- network
- packet
- bandwidth
- grouping
- priority
- 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
Abstract
The present invention discloses an acceleration method of an asymmetric network or a concurrent network. In the method, firstly bearable bandwidth of the network and the actual processing bandwidth are estimated; when the packet loss or the packet retransmission occurs, if the bearable bandwidth of the network is larger than the actual processing bandwidth, the cause is considered to be the low packet processing of the sending terminal and the receiving terminal, and a lower priority is given to the confirm (ACK) packet. When the packet loss or the packet retransmission occurs, if the bearable bandwidth of the network is smaller than the actual processing bandwidth, the cause is considered to be the network congestion, then the packet is given and confirmed a higher priority; and the transmission rate of the confirm frame is adjusted according to the priority grade of the confirm packet. The present invention opens up different grades of green channels for the confirm packet, to avoid the confirm packet-intensity phenomenon, and the network congestion problem caused by the confirm packet-intensity is fundamentally solved, and the network security hidden trouble caused by losing the control of the confirm packet is avoided.
Description
Technical field
The present invention relates to a kind of in communication network the method for speeding up data transmission, relate in particular to a kind of characteristics at asymmetric and how concurrent TCP/IP network application, the transfer of data accelerated method that the affirmation of flexible utilization Transmission Control Protocol (ACK) grouping realizes belongs to network congestion control (congestioncontrol) technical field.
Background technology
At present, transmission control protocol/Internet Protocol (TCP/IP) has become widely used standard communication protocol.Wherein, Internet Protocol (IP) is used to provide the service of submitting of " packet " as network layer protocol, and transmission control protocol (TCP) is submitted the connection-oriented transport layer services of structure on the service basis as transport layer protocol at packet.
What adopt based on the internet of ICP/IP protocol is connectionless end-to-end packet exchange, and " doing one's best " (best effort) is provided the design mechanism of service model.Along with the rapid expansion of number of the Internet users, the message amount that needs transmission is also in rapid increase.When having too much message in the network, the performance of network can descend, and this phenomenon is called congested.When network congestion occurring, because formation such as overflows at the influence of factor, router can abandon about 10% packet, thereby the reliability that logarithm reportedly send is brought adverse influence.
Congested control is exactly that network node takes measures to avoid congested generation or to congested making a response.According to statistics, what 95% data traffic was used on the internet at present is ICP/IP protocol, therefore, provides special congestion control mechanism the Control Network congestion problems to be had the meaning of particular importance at ICP/IP protocol.
Relevant studies show that, the reliability that network data transmits mainly is subjected to three factor affecting: the rearrangement of data corruption, loss of data and data.In existing Transmission Control Protocol, come management data to lose by timeout mechanism; By when section arrives the recipient, they being carried out verification and come management data to worsen; By confirming that (ACK) grouping waits the rearrangement of management data.Owing to timeout mechanism is also closely related with affirmation (ACK) grouping, confirm that therefore (ACK) transmission packets situation will losing and the generation great influence of resequencing network data.For Transmission Control Protocol, confirm that abandoning with congested of (ACK) grouping means that all relevant data segments will retransmit, therefore, by confirming that the network congestion that (ACK) grouping causes has become a principal element that causes that the whole network is congested.
At present, in the patent No. is in the Chinese invention patent " a kind of adaptive congestion control method of suitable wire/wireless hybrid network " of ZL 03155730.9, judge the congestion condition of link according to the relative one-way delay variation tendency of confirming (ACK) grouping, and then distinguish packet loss be because link congestion cause or because wireless makeing mistakes causes, take different Restoration Mechanism according to different packet loss reasons adaptively.In addition, be in the Chinese invention patent application of 200610005010.X at application number, the transmission control protocol jamming control method that uses a plurality of transmission control protocols to confirm is provided.This method comprises: when having packet loss in the grouping that mobile node is receiving, receive the grouping of retransmitting from corresponding node; The number of the affirmation message that the calculating mobile node will send; Produce a plurality of acknowledge messages according to calculated number, and a plurality of acknowledge messages are sent to sending node; And the corresponding congestion window value of a plurality of acknowledge messages that receives of increase and sending node, and carry out TCP and send.
In general, existing TCP congestion control mechanism mainly is a transmitting-receiving frequency of regulating the transmit leg data message by the transmitting-receiving situation of confirming (ACK) grouping itself, is taked different control methods to avoid or alleviating network congestion by data receiver according to the situation of received affirmation (ACK) grouping.But, be that the asymmetric and multi-concurrency network of representative makes existing network congestion controlling mechanism be in runaway condition with interactive peer-to-peer communications, its reason mainly contains following 3 points: 1. existing TCP congestion control mechanism is reactive poor in express network; 2. network asymmetry (NetworkAsymmetry) is the main cause that causes network congestion, the existing method of improving the TCP performance remains First Come First Served under the principle of (Best Effort) of doing one's best (FirstIn First Serve) for different application, and the problem that the network asymmetry is caused does not obtain substantial solution; 3.TCP congestion control mechanism mainly is to be the static method of measure at the source end with " prevention " and " recovery ", to the trace ability of transfer of data and poor to the adaptive capacity of heterogeneous networks application data attribute change.Therefore, existing congested control technology does not fully take into account technical characterstic asymmetric and how concurrent in the internet, and is little to solving the network congestion problem effect that causes because of asymmetric and how concurrent characteristics.
Summary of the invention
In view of the deficiencies in the prior art, the purpose of this invention is to provide a kind of accelerated method of asymmetric and multi-concurrency network.Affirmation (ACK) grouping in this method flexible utilization Transmission Control Protocol can effectively improve based on the Transmission Control Protocol of using and transmit congested control performance, improves asymmetric and transmission speed multi-concurrency network, reduces the probability that network congestion takes place.
For realizing above-mentioned goal of the invention, the present invention adopts following technical scheme:
A kind of accelerated method of asymmetric and multi-concurrency network is taked the dynamic management of differentiation on based on the basis of traffic classification to affirmation (ACK) grouping of different application, it is characterized in that:
Estimate that at first network can bear bandwidth and actual processing bandwidth;
If when packet loss or packet retransmissions present took place, network can bear bandwidth greater than actual processing bandwidth, think that then packet loss or packet retransmissions present are slowly caused by transmitting terminal and receiving terminal processing grouping, give affirmation (ACK) grouping lower priority;
When if packet loss or packet retransmissions present take place, network can bear bandwidth greater than network less than actual processing bandwidth, think that then packet loss or packet retransmissions present are caused by network congestion, give affirmations (ACK) higher priority that divides into groups;
Confirm the transfer rate of (ACK) grouping according to the priority adjustment of described affirmation (ACK) grouping.
Wherein, at first setting up a TCP connects, and bandwidth can be born and actual processing bandwidth is sampled to the network of this connection, and can bear the bandwidth sampled value and the actual processing bandwidth sampled value is carried out smoothing processing to the network that repeatedly obtains, obtain the estimated value that current network can be born bandwidth and actual processing bandwidth.
When if packet loss or packet retransmissions present take place, network can bear bandwidth greater than actual processing bandwidth and default threshold value sum, think that then packet loss or packet retransmissions present are slowly caused by transmitting terminal and receiving terminal processing grouping, give described affirmation (ACK) grouping first priority.
When if packet loss or packet retransmissions present take place, network can bear bandwidth poor less than actual processing bandwidth and default threshold value, thinks that then packet loss or packet retransmissions present are caused by network congestion, gives described affirmation (ACK) grouping second priority.
When if packet loss or packet retransmissions present take place, network can bear bandwidth smaller or equal to actual processing bandwidth and default threshold value sum, and poor more than or equal to actual processing bandwidth and default threshold value, think that then packet loss or packet retransmissions present are to be caused by the other reasons beyond network congestion and transmitting terminal and the receiving terminal processing grouping slowly, give described affirmation (ACK) grouping the 3rd priority.
Described threshold value is initially (64*8)/RTT, and wherein RTT is the two-way time of grouping.
Described the 3rd priority is greater than described second priority, and described second priority is greater than described first priority.
When packet loss or packet retransmissions present occurring, carry out the adjustment of congestion window and slow enabling gate limit value according to congestion control policy in the TCP-Reno algorithm.
Dispatch the grouping that described affirmation (ACK) is divided into groups according to the height of described priority, guarantee the transfer rate of affirmation (ACK) grouping that priority is higher.
When affirmation (ACK) grouping that handle to send, always check in the formation with higher priority whether have the grouping message at first, if grouping is arranged then, directly send with this packet scheduling dequeue; Have only when it is empty queue, go the lower formation of dispatching priority again.
Network accelerating method provided by the present invention is distinguished affirmation (ACK) grouping and the data message for the treatment of the whole network according to user's real needs, affirmation (ACK) grouping of different application is treated in differentiation, and open up the green channel of different stage based on being applied as affirmation (ACK) grouping, make and confirm that the intensive phenomenon of (ACK) grouping no longer takes place, fundamentally solved by confirming the intensive network congestion problem of bringing of (ACK) grouping, avoided by confirming that (ACK) divides into groups out of control and network security hidden danger that cause.
Description of drawings
The present invention is further illustrated below in conjunction with the drawings and specific embodiments.
Fig. 1 has shown the basic structure of the data handling system network that is used to implement the method for the invention;
Fig. 2 is Peer to Peer (abbreviating P2P as) the application architecture figure as typical asymmetric and how concurrent application;
Fig. 3 is the example of 4 layers of communication construction of employing ICP/IP protocol;
Fig. 4 is a Transmission Control Protocol its working principles schematic diagram;
Fig. 5 is a structural representation of confirming (ACK) grouping in the Transmission Control Protocol;
Fig. 6 is the implementation step flow chart of the accelerated method of asymmetric and multi-concurrency network of the present invention.
Embodiment
Following with reference to accompanying drawing and specifically with reference to Fig. 1, the data handling system network that wherein can be used for realizing the preferred embodiments of the present invention is described.Data handling system network 100 logically is divided into transport plane 101, service plane 102 and using planar 103.Wherein, transport plane 101 comprises core transport layer 104, Access Layer 105, fringe node 106 etc.The data of this data handling system network 100 transmit follows the TCP/IP standard.
Fig. 2 is Peer to Peer (abbreviating P2P as) the application architecture figure as typical asymmetric and how concurrent application.The P2P technology has been broken traditional client-server model, makes the network user can be directly connected to other user's computer, swap file, rather than is connected to server as before and goes to browse and download.All elements is in the status of complete equity in this framework, does not have the branch of client computer, server, and promptly each node is a client computer, is again server, both provides service to others, also obtains service there from others.
Fig. 3 has shown 4 layers of communication construction 300 that are used to carry out ICP/IP protocol, and this communication construction comprises application layer 310, transmitting layer 3 12, network layer 314 and link layer 316.Wherein, link layer 316 (being also referred to as data link layer or network interface layer) generally includes network interface unit corresponding in device driver in the operating system and the computer.They handle all hardware details of docking with the network medium that is using (for example Ethernet cable etc.) physics together.
Network layer 314 (being also referred to as the Internet layer) is handled moving of network packet everywhere.For example, the route of the various packets that on network, transmit of network layer handles.Network layer in the Transmission Control Protocol group is made up of several agreements, comprises IP (Internet Protocol), ICMP (Internet Control Message Protocol) and ICMP (Internet Group Management Protocol).
Transmitting layer 3 12 provides the interface between network layer 314 and the application layer 310, and it helps the transmission of data between two master computers.The data that transport layer is responsible for passing to it from application layer are divided into the data block (chunk) of the suitable size that is suitable for network layer, grouping that affirmation is received or the like.In the ICP/IP protocol group, there are two kinds of host-host protocol: TCP (transmission control protocol) and UDP (User Datagram Protocol).TCP provides reliability services, and to guarantee correctly to transmit data between two main frames, this service comprises loss detection and the service of retransferring.On the contrary, UDP only sends to another by the packet that will be called packet from a main frame, does not provide the user to guarantee correctly to transmit any mechanism of data.
When application layer 310 uses ICP/IP protocol to send data, at first with data by sending to protocol stack under each course, every layer by adding information on the data on the data that header appended in advance its reception.For example, application header is appended on the user data in advance, to form application data in application layer.In transport layer, transport protocol header is appended on the application data in advance.If transport layer is TCP, the TCP header is appended on the application data in advance, form the TCP frame thus, it is sent to network layer IP.The TCP header comprises 20 bytes.Similarly, in network layer, network layer header is appended on the transport layer data in advance.Under the situation of TCP/IP, the IP header is appended to the TCP frame in advance, to form the IP packet.The IP header also comprises 20 bytes.At last, at link layer, the media header of for example ethernet header is added on the data of network layer reception, to form Frame.
Fig. 4 has shown the basic functional principle of TCP.Transmission Control Protocol is connection-oriented.So-called when connection is meant the computer intercommunication, must at first connect, and then transmit data.Transmission Control Protocol is taken a three-stage approach when connecting: the first step is that request end (client) sends a TCP message that comprises SYN synchronous (Synchronize) sign, and the SYN sync message can indicate the port of client use and the initial sequence number that TCP connects; In second step, server will return the message of a SYN+ACK after receiving the SYN message of client, and the request of expression client is accepted, and the TCP sequence number is added one simultaneously, and ACK promptly confirms; In the 3rd step, client is also returned a confirmation message ACK and is given server end, and same TCP sequence number is added one, connects to this TCP and finishes.
Fig. 5 is a frame structure schematic diagram of confirming (ACK) grouping in the Transmission Control Protocol.TCP acknowledgment (ACK) grouping 522 comprises media header 524, protocol header 526, the sequence-number field 528 that is received and sends row field 530 etc.Be responsible for reception data computing machine and confirm (ACK) grouping (522), can confirm that transmission or retry message receive by sending.
Some background knowledge of Transmission Control Protocol and affirmation (ACK) grouping have been introduced above.Introduce the transfer of data accelerated method that utilizes affirmation (ACK) grouping in the Transmission Control Protocol to realize provided by the present invention below in detail.
The core concept of this method is on based on the basis of traffic classification differentiation dynamic management mechanism to be taked in affirmation (ACK) grouping of different application, promptly by network delay, reception and the transmission lag that divides into groups in the computing network, the RTT (two-way time) of grouping and the speed of affirmation (ACK) grouping, determine whether that according to network delay, RTT and ACK speed packet loss takes place then or do not have retransmission packet and the reason of generation network congestion, and adopt different strategies to carry out priority scheduling and speed adjustment confirming (ACK) grouping at different reasons.
In the present invention, utilize RTT (two-way time) the estimation network of network delay and grouping can bear bandwidth BW A (Band Width Acceptable), utilized the speed of affirmation (ACK) grouping and the actual processing bandwidth BWR (BandWidth Real) of average mark group length estimation network transmitting terminal and receiving terminal simultaneously.When estimation after network can bear bandwidth and actual processing bandwidth, just can bear bandwidth BW A and actual processing bandwidth BWR judges network congestion degree and network congestion reason according to network.If when packet loss or packet retransmissions present took place, network can bear bandwidth BW A greater than actual processing bandwidth BWR, think that then packet loss or packet retransmissions present are slowly caused by transmitting terminal and receiving terminal processing grouping; When if packet loss or packet retransmissions present take place, network can bear bandwidth greater than network B WA less than actual processing bandwidth BWR, think that then packet loss or packet retransmissions present are caused by network congestion.At packet loss that causes based on different reasons and packet retransmissions present, just can take Different Strategies to confirming the adjustment of (ACK) packet-priority queue scheduling and affirmation (ACK) packet rates.
Below in conjunction with Fig. 6, and be further described according to the RTT (two-way time) of network delay, grouping and the rate calculations process of confirming (ACK) concrete implementation step to this method.
Step 1: determine that according to the IP five-tuple information of grouping TCP connects, and to the network of this connection can bear bandwidth BW A sample greater than actual processing bandwidth BWR obtain Sample_BWA[k] and Sample_BWR[k] sampled value.
Sample_BWA[k]＝(Pkt_size_sum*8)/(ACKed_time-last_PSH_ACK_time)
Sample_BWR[k]＝(wnd*8)/(ACKed_time-last_ACK_time)
Wherein, IP five-tuple information comprises: source IP address, purpose IP address, source port number, destination slogan, protocol number.
Pkt_size_sum is illustrated in the message length of network node C place this connection of statistics, and unit is a byte.
ACKed_time is illustrated in the time that network node C place this connection confirmation message of statistics arrives, and unit is second.
Last_PSH_ACK_time is illustrated in the time of last data message (the TCP message that comprises psh and the ack sign) arrival of network node C place this connection of statistics, and unit is second.
Wnd represents that receiving terminal wishes the byte number that receives.
K represents the k time sampling, k=1,2,3 ...
Step 2: can bear bandwidth BW A[k to network] sampled value and actual processing bandwidth BWR[k] sampled value carries out smoothing processing, obtains the estimated value that current network can be born bandwidth BW A and actual processing bandwidth BWR.
BWA[k]＝(19/21)*BWA[k-1]+(1/21)*(Sample_BWA[k]+Sample_
BWA[k-1])
BWR[k]＝(19/21)*BWR[k-1]+(1/21)*(Sample_BWR[k]+Sample_BWR[k-1])
Step 3: if when packet loss or packet retransmissions present take place, network can bear bandwidth BW A greater than actual processing bandwidth BWR and threshold value B eta sum, think that then packet loss or packet retransmissions present are slowly caused by transmitting terminal and receiving terminal processing grouping, execution in step 4, otherwise the content below the continuation execution in step 3.
When if packet loss or packet retransmissions present take place, network can bear bandwidth BW A poor less than actual processing bandwidth BWR and threshold value B eta, think that then packet loss or packet retransmissions present are caused by network congestion, then execution in step 5, otherwise continue the content below the execution in step 3.
When if packet loss or packet retransmissions present take place, network can bear bandwidth BW A smaller or equal to actual processing bandwidth BWR and threshold value B eta sum, and poor more than or equal to actual processing bandwidth BWR and threshold value B eta, be BWR-Beta=＜BWA=＜BWR+Beta, think that then packet loss or packet retransmissions present are to be caused by the other reasons beyond network congestion and transmitting terminal and the receiving terminal processing grouping slowly, then execution in step 6, otherwise continue the content below the execution in step 3.
Described threshold value B eta is an empirical value normally, can obtain threshold value B eta=(64*8)/RTT when initial by the real network test.
The calculating of RTT can be calculated according to the method among the RFC2681.When existing the burst service or the burst network equipment unusual in the network, the value of RTT can change, and the RTT value in this method is the real-time change value, and the use of beta value and RTT certainty of measurement have direct relation, is generally millisecond (ms) class precision.The Beta value is generally 1kbps～10kbps, and specifically the class of business by network and the transmission of network kind determines.
Step 4: priority P=1 of confirming (ACK) grouping, at this moment, think network bandwidth unsaturation, producing packet loss or packet retransmissions present is slowly caused by transmitting terminal and receiving terminal processing grouping, and (detailed adjustment process can be consulted " TCP/IP detailed annotation volume 1: agreement " that China Machine Press publishes in April, 2000 (ISBN:7111075668) the 21st chapter in the book to carry out the adjustment of congestion window and slow enabling gate limit value according to congestion control policy in the algorithm of existing TCP-Reno, especially the relevant explanation in the 21.6 joint congestion avoidance algorithms will not be given unnecessary details herein); Jump procedure 7.
Step 5: priority P=2 of confirming (ACK) grouping, at this moment, think network bandwidth unsaturation, produce packet loss or packet retransmissions present and slowly cause, and carry out the adjustment of congestion window and slow enabling gate limit value according to congestion control policy in the algorithm of existing TCP-Reno by transmitting terminal and receiving terminal processing grouping; Jump procedure 7.
Step 6: priority P=3 of confirming (ACK) grouping, at this moment, think network bandwidth unsaturation, produce packet loss or packet retransmissions present and slowly cause, and carry out the adjustment of congestion window and slow enabling gate limit value according to congestion control policy in the algorithm of existing TCP-Reno by transmitting terminal and receiving terminal processing grouping; Jump procedure 7.
Step 7: (ACK) grouping is confirmed in the size scheduling according to the priority P value, guarantees big affirmation (ACK) packet rates of priority P value.When having the grouping of high priority, at first send the grouping in priority P=3 formations, after the transmission of the grouping in the high-priority queue disposes, in the grouping that sends priority P=2 and the grouping of priority P=3.When handle sending grouping, always check in the formation of priority P=3 whether have the grouping message at first, if grouping is arranged then, directly send with this packet scheduling dequeue; When the formation of priority P=3 is empty queue, the formation of removing dispatching priority P=2 and P=1 again.
Just to be defined as the big more priority of numerical value high more for priority herein, the priority when priority was greater than P=1 when priority was greater than P=2 during as P=3.
Above-mentioned steps 1～step 4 has been described network can bear the acquisition process of bandwidth BW A less than actual processing bandwidth BWR, and the method for phase-split network packet loss and re-transmission and the method for affirmation (ACK) packet scheduling have been described in step 5～8.By the specific embodiment of the invention described above as can be seen, the present invention has not only kept the employing congestion window of TCP-Reno and has adjusted the algorithm advantage of handling network congestion and packet loss re-transmission, increased method of bearing bandwidth BW A Network Based simultaneously less than affirmation (ACK) the packet-priority scheduling of actual processing bandwidth BWR analysis, handle the network congestion of different reasons, effectively utilize the network bandwidth and estimated scheduling affirmation (ACK) grouping, network bandwidth utilance is improved, improve the fairness that network uses simultaneously, prevented the generation of network congestion and periods of network disruption.
Describe the accelerated method of asymmetric and multi-concurrency network provided by the present invention above in detail.For one of ordinary skill in the art, any conspicuous change of under the prerequisite that does not deviate from connotation of the present invention it being done all will constitute to infringement of patent right of the present invention, with corresponding legal responsibilities.
Claims (10)
1. the accelerated method of an asymmetric and multi-concurrency network is taked the dynamic management of differentiation on based on the basis of traffic classification to the affirmation grouping of different application, it is characterized in that:
Estimate that at first network can bear bandwidth and actual processing bandwidth;
If when packet loss or packet retransmissions present took place, network can bear bandwidth greater than actual processing bandwidth, think that then packet loss or packet retransmissions present are slowly caused by transmitting terminal and receiving terminal processing grouping, give and confirm the grouping lower priority;
When if packet loss or packet retransmissions present take place, network can bear bandwidth greater than network less than actual processing bandwidth, think that then packet loss or packet retransmissions present are caused by network congestion, give and confirm the grouping higher priority;
Confirm the transfer rate of grouping according to the priority adjustment of described affirmation grouping.
2. the accelerated method of asymmetric and multi-concurrency network as claimed in claim 1 is characterized in that:
At first setting up a TCP connects, and bandwidth can be born and actual processing bandwidth is sampled to the network of this connection, and can bear the bandwidth sampled value and the actual processing bandwidth sampled value is carried out smoothing processing to the network that repeatedly obtains, obtain the estimated value that current network can be born bandwidth and actual processing bandwidth.
3. the accelerated method of asymmetric and multi-concurrency network as claimed in claim 1 is characterized in that:
When if packet loss or packet retransmissions present take place, network can bear bandwidth greater than actual processing bandwidth and default threshold value sum, think that then packet loss or packet retransmissions present handle grouping by transmitting terminal and receiving terminal and slowly cause, give described affirmation first priority of dividing into groups.
4. the accelerated method of asymmetric and multi-concurrency network as claimed in claim 1 is characterized in that:
When if packet loss or packet retransmissions present take place, network can bear bandwidth poor less than actual processing bandwidth and default threshold value, thinks that then packet loss or packet retransmissions present are caused by network congestion, gives described affirmation second priority of dividing into groups.
5. the accelerated method of asymmetric and multi-concurrency network as claimed in claim 1 is characterized in that:
When if packet loss or packet retransmissions present take place, network can bear bandwidth smaller or equal to actual processing bandwidth and default threshold value sum, and poor more than or equal to actual processing bandwidth and default threshold value, think that then packet loss or packet retransmissions present are to be caused by the other reasons that network congestion and transmitting terminal and receiving terminal are handled beyond the grouping slowly, give described affirmation the 3rd priority of dividing into groups.
6. as the accelerated method of any described asymmetric and multi-concurrency network in the claim 3～5, it is characterized in that:
Described threshold value is initially 64*8/RTT, and wherein RTT is the two-way time of grouping.
7. as the accelerated method of any described asymmetric and multi-concurrency network in the claim 3～5, it is characterized in that:
Described the 3rd priority is greater than described second priority, and described second priority is greater than described first priority.
8. as the accelerated method of any described asymmetric and multi-concurrency network in the claim 3～5, it is characterized in that:
When packet loss or packet retransmissions present occurring, carry out the adjustment of congestion window and slow enabling gate limit value according to congestion control policy in the TCP-Reno algorithm.
9. as the accelerated method of any described asymmetric and multi-concurrency network in the claim 3～5, it is characterized in that:
Dispatch described affirmation grouping according to the height of described priority, guarantee the transfer rate of the affirmation grouping that priority is higher.
10. the accelerated method of asymmetric and multi-concurrency network as claimed in claim 9 is characterized in that:
When the affirmation grouping that handle to send, always check in the formation with higher priority whether have the grouping message at first, if grouping is arranged then, directly send with this packet scheduling dequeue; Have only when it is empty queue, go the lower formation of dispatching priority again.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2007101111482A CN100534069C (en) | 2007-04-26 | 2007-06-14 | Accelerating method for asymmetric and multi-concurrency network |
Applications Claiming Priority (3)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200710098772.3 | 2007-04-26 | ||
| CN200710098772 | 2007-04-26 | ||
| CNB2007101111482A CN100534069C (en) | 2007-04-26 | 2007-06-14 | Accelerating method for asymmetric and multi-concurrency network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101110767A true CN101110767A (en) | 2008-01-23 |
| CN100534069C CN100534069C (en) | 2009-08-26 |
Family
ID=39042665
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB2007101111482A Expired - Fee Related CN100534069C (en) | 2007-04-26 | 2007-06-14 | Accelerating method for asymmetric and multi-concurrency network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN100534069C (en) |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101848130A (en) * | 2009-03-27 | 2010-09-29 | 北京大学深圳研究生院 | Method based on response drive for controlling UDP congestion of P2P streaming media |
| CN102420684A (en) * | 2011-11-04 | 2012-04-18 | 展讯通信（上海）有限公司 | Processing method of TCP (Transmission Control Protocol) data stream and communication device |
| CN102611653A (en) * | 2012-03-04 | 2012-07-25 | 黄东 | Bandwidth estimation method based on hybrid network |
| WO2013167024A2 (en) * | 2012-10-25 | 2013-11-14 | 中兴通讯股份有限公司 | Method and apparatus for controlling tcp bidirectional concurrent transmission |
| CN105847183A (en) * | 2016-03-28 | 2016-08-10 | 乐视控股（北京）有限公司 | Bandwidth distribution method of content distribution network and bandwidth distribution system of content distribution network |
| CN106341348A (en) * | 2015-07-17 | 2017-01-18 | 普天信息技术有限公司 | TCP service oriented flow control method and access network element |
| CN106850169A (en) * | 2016-11-29 | 2017-06-13 | 上海华为技术有限公司 | The method and communication equipment of a kind of scheduling of resource distribution |
| CN109495388A (en) * | 2018-11-30 | 2019-03-19 | 四川安迪科技实业有限公司 | TCP accelerated method based on satellite communication |
| CN109525374A (en) * | 2017-09-20 | 2019-03-26 | 华为技术有限公司 | Method, wireless access point, user equipment and the transmission device of data transmission |
| CN110351114A (en) * | 2019-05-23 | 2019-10-18 | 蓝炬兴业(赤壁)科技有限公司 | Based on industrial personal computer secure network accelerated method and system |
| CN111294547A (en) * | 2020-02-20 | 2020-06-16 | 江苏齐德隆智能科技有限公司 | Remote online inquiry and evidence obtaining system and method |
| CN116366553A (en) * | 2023-03-08 | 2023-06-30 | 杭州粒子光速科技有限公司 | Self-adaptive network congestion control method |
-
2007
- 2007-06-14 CN CNB2007101111482A patent/CN100534069C/en not_active Expired - Fee Related
Cited By (23)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101848130A (en) * | 2009-03-27 | 2010-09-29 | 北京大学深圳研究生院 | Method based on response drive for controlling UDP congestion of P2P streaming media |
| CN102420684A (en) * | 2011-11-04 | 2012-04-18 | 展讯通信（上海）有限公司 | Processing method of TCP (Transmission Control Protocol) data stream and communication device |
| CN102420684B (en) * | 2011-11-04 | 2014-11-05 | 展讯通信（上海）有限公司 | Processing method of TCP (Transmission Control Protocol) data stream and communication device |
| CN102611653B (en) * | 2012-03-04 | 2014-09-24 | 黄东 | Bandwidth estimation method based on hybrid network |
| CN102611653A (en) * | 2012-03-04 | 2012-07-25 | 黄东 | Bandwidth estimation method based on hybrid network |
| CN103780336B (en) * | 2012-10-25 | 2017-07-07 | 中兴通讯股份有限公司 | The two-way concurrent transmission regulation and control method and apparatus of TCP |
| WO2013167024A2 (en) * | 2012-10-25 | 2013-11-14 | 中兴通讯股份有限公司 | Method and apparatus for controlling tcp bidirectional concurrent transmission |
| WO2013167024A3 (en) * | 2012-10-25 | 2014-01-03 | 中兴通讯股份有限公司 | Method and apparatus for controlling tcp bidirectional concurrent transmission |
| CN106341348B (en) * | 2015-07-17 | 2019-06-21 | 普天信息技术有限公司 | A kind of flow control methods and access network elements towards TCP traffic |
| CN106341348A (en) * | 2015-07-17 | 2017-01-18 | 普天信息技术有限公司 | TCP service oriented flow control method and access network element |
| CN105847183A (en) * | 2016-03-28 | 2016-08-10 | 乐视控股（北京）有限公司 | Bandwidth distribution method of content distribution network and bandwidth distribution system of content distribution network |
| CN106850169B (en) * | 2016-11-29 | 2019-11-29 | 上海华为技术有限公司 | A kind of method and communication equipment of scheduling of resource distribution |
| WO2018099201A1 (en) * | 2016-11-29 | 2018-06-07 | 华为技术有限公司 | Method for scheduling and allocating resources, and communication device |
| CN106850169A (en) * | 2016-11-29 | 2017-06-13 | 上海华为技术有限公司 | The method and communication equipment of a kind of scheduling of resource distribution |
| CN109525374A (en) * | 2017-09-20 | 2019-03-26 | 华为技术有限公司 | Method, wireless access point, user equipment and the transmission device of data transmission |
| CN109525374B (en) * | 2017-09-20 | 2020-10-16 | 华为技术有限公司 | Data transmission method, wireless access point, user equipment and transmission equipment |
| CN109495388A (en) * | 2018-11-30 | 2019-03-19 | 四川安迪科技实业有限公司 | TCP accelerated method based on satellite communication |
| CN109495388B (en) * | 2018-11-30 | 2021-04-27 | 四川安迪科技实业有限公司 | TCP acceleration method based on satellite communication |
| CN110351114A (en) * | 2019-05-23 | 2019-10-18 | 蓝炬兴业(赤壁)科技有限公司 | Based on industrial personal computer secure network accelerated method and system |
| CN110351114B (en) * | 2019-05-23 | 2023-03-28 | 蓝炬兴业(赤壁)科技有限公司 | Industrial computer based secure network acceleration method and system |
| CN111294547A (en) * | 2020-02-20 | 2020-06-16 | 江苏齐德隆智能科技有限公司 | Remote online inquiry and evidence obtaining system and method |
| CN116366553A (en) * | 2023-03-08 | 2023-06-30 | 杭州粒子光速科技有限公司 | Self-adaptive network congestion control method |
| CN116366553B (en) * | 2023-03-08 | 2023-11-10 | 杭州流形新网络科技有限公司 | Self-adaptive network congestion control method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN100534069C (en) | 2009-08-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN100534069C (en) | Accelerating method for asymmetric and multi-concurrency network | |
| Balakrishnan et al. | How network asymmetry affects TCP | |
| EP1295428B1 (en) | Performance enhancement of transmission control protocol (tcp) for wireless network applications | |
| US9467390B2 (en) | Method and device for data transmission | |
| US20060165029A1 (en) | Protecting real-time data in wireless networks | |
| JP2001268130A (en) | Method and device for controlling flow of tcp connection on flow-controlled network, and interconnection equipment | |
| CN102474455B (en) | Network traffic accelerator | |
| Abed et al. | Architecture And Functional Structure Of Transmission Control Protocol Over Various Networks Applications | |
| CN103346963A (en) | MPTCP data scheduling method based on forecast arrival time | |
| CN102801692B (en) | A kind of transmission control protocol optimization method and system connecting based on division | |
| Ahmad et al. | Enhancing fast TCP’s performance using single TCP connection for parallel traffic flows to prevent head-of-line blocking | |
| De Figueiredo et al. | Using coloured Petri nets to investigate behavioural and performance issues of TCP protocols | |
| Henderson | TCP performance over satellite channels | |
| CN100563199C (en) | A kind of method that in wireless network, improves a plurality of Non-Access Stratum switching performances | |
| Yadav et al. | A review of congestion control mechanisms for wireless networks | |
| Elloumi et al. | A simulation-based study of TCP dynamics over HFC networks | |
| Hurtig et al. | SCTP: designed for timely message delivery? | |
| Chen et al. | An improved rudp for data transmission in embedded real-time system | |
| Heuschkel et al. | Udp++: Cross-layer optimizable transport protocol for managed wireless networks | |
| Wang et al. | Reliable and efficient data transfer protocol based on UDP in cluster system | |
| Jiapei et al. | Performance simulation analysis of transport layer protocol for satellite network | |
| Zhang et al. | Performance simulation analysis of transport layer protocol for satellite network | |
| CN115766747A (en) | Reliable and efficient transmission method suitable for wireless channel | |
| Radovanovic et al. | Improving TCP/IP performance over last-hop wireless networks for streaming video delivery | |
| Foerster et al. | Reducing the latency-tail of short-lived flows: Adding forward error correction in data centers |
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 | ||
| ASS | Succession or assignment of patent right |
Owner name: HEFEI HOT INFORMATION SCIENCE AND TECHNOLOGY CO., Free format text: FORMER OWNER: BEIJING QQ TECHNOLOGY CO., LTD. Effective date: 20100601 |
|
| C41 | Transfer of patent application or patent right or utility model | ||
| COR | Change of bibliographic data |
Free format text: CORRECT: ADDRESS; FROM: 100037 8/F, BLOCK B, WANTONG XINSHIJIE PLAZA, NO.2, FUWAI STREET, XICHENG DISTRICT, BEIJING TO: 230088 ROOM 320, MINCHUANG CENTER, NO.605, HUANGSHAN ROAD, HIGH-TECH. ZONE, HEFEI CITY, ANHUI PROVINCE |
|
| TR01 | Transfer of patent right |
Effective date of registration: 20100601 Address after: 320 room 230088, center of 605 people's road, Mount Huangshan Road, hi tech Zone, Anhui, Hefei Patentee after: Hefei Haote Information Technology Co., Ltd. Address before: 100037 Beijing City, Xicheng District Fuwai Street No. 2 Wantong New World Plaza B block 8 layer Patentee before: Beijing QQ Technology Co., Ltd. |
|
| ASS | Succession or assignment of patent right |
Owner name: BEIJING QQ TECHNOLOGY CO., LTD. Free format text: FORMER OWNER: HEFEI HAOTE INFORMATION TECHNOLOGY CO., LTD. Effective date: 20110701 |
|
| C41 | Transfer of patent application or patent right or utility model | ||
| COR | Change of bibliographic data |
Free format text: CORRECT: ADDRESS; FROM: 230088 ROOM 320, INNOVATION CENTER, 605, HUANGSHAN ROAD, HIGH-TECH. ZONE, HEFEI CITY, ANHUI PROVINCE TO: 100037 8/F, TOWER B, WANTONG NEW WORLD SQUARE, NO. 2, FUWAI STREET, XICHENG DISTRICT, BEIJING |
|
| TR01 | Transfer of patent right |
Effective date of registration: 20110701 Address after: 100037 Beijing City, Xicheng District Fuwai Street No. 2 Wantong New World Plaza B block 8 layer Patentee after: Beijing QQ Technology Co., Ltd. Address before: 320 room 230088, center of 605 people's road, Mount Huangshan Road, hi tech Zone, Anhui, Hefei Patentee before: Hefei Haote Information Technology Co., Ltd. |
|
| ASS | Succession or assignment of patent right |
Owner name: BEIJING CHANGXUN HENGXING NETWORK TECHNOLOGY CO., Free format text: FORMER OWNER: BEIJING QQ TECHNOLOGY CO., LTD. Effective date: 20121206 |
|
| C41 | Transfer of patent application or patent right or utility model | ||
| COR | Change of bibliographic data |
Free format text: CORRECT: ADDRESS; FROM: 100037 XICHENG, BEIJING TO: 100036 HAIDIAN, BEIJING |
|
| TR01 | Transfer of patent right |
Effective date of registration: 20121206 Address after: 100036, room 134, building A, Ho Ho building, No. 3, north depression road, Beijing, Haidian District Patentee after: Beijing Chang Xing Star Network Technology Co., Ltd. Address before: 100037 Beijing City, Xicheng District Fuwai Street No. 2 Wantong New World Plaza B block 8 layer Patentee before: Beijing QQ Technology Co., Ltd. |
|
| ASS | Succession or assignment of patent right |
Owner name: HEFEI HAOTE INFORMATION TECHNOLOGY CO., LTD. Free format text: FORMER OWNER: BEIJING CHANGXUN HENGXING NETWORK TECHNOLOGY CO., LTD. Effective date: 20130106 |
|
| C41 | Transfer of patent application or patent right or utility model | ||
| COR | Change of bibliographic data |
Free format text: CORRECT: ADDRESS; FROM: 100036 HAIDIAN, BEIJING TO: 230088 HEFEI, ANHUI PROVINCE |
|
| TR01 | Transfer of patent right |
Effective date of registration: 20130106 Address after: 320 room 605, min Chong Center, 230088 Mount Huangshan Road, Hefei high tech Zone, Anhui, China Patentee after: Hefei Haote Information Technology Co., Ltd. Address before: 100036, room 134, building A, Ho Ho building, No. 3, north depression road, Beijing, Haidian District Patentee before: Beijing Chang Xing Star Network Technology Co., Ltd. |
|
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20090826 Termination date: 20150614 |
|
| EXPY | Termination of patent right or utility model |