CN105634875A - Method and system for identifying packet loss type in data transmission of reliable transmission protocol - Google Patents
Method and system for identifying packet loss type in data transmission of reliable transmission protocol Download PDFInfo
- Publication number
- CN105634875A CN105634875A CN201610082255.6A CN201610082255A CN105634875A CN 105634875 A CN105634875 A CN 105634875A CN 201610082255 A CN201610082255 A CN 201610082255A CN 105634875 A CN105634875 A CN 105634875A
- Authority
- CN
- China
- Prior art keywords
- packet
- loss
- rtt
- packet loss
- base
- 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.)
- Pending
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L43/00—Arrangements for monitoring or testing data switching networks
- H04L43/08—Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters
- H04L43/0823—Errors, e.g. transmission errors
- H04L43/0829—Packet loss
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L47/00—Traffic control in data switching networks
- H04L47/10—Flow control; Congestion control
- H04L47/28—Flow control; Congestion control in relation to timing considerations
- H04L47/283—Flow control; Congestion control in relation to timing considerations in response to processing delays, e.g. caused by jitter or round trip time [RTT]
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L69/00—Network arrangements, protocols or services independent of the application payload and not provided for in the other groups of this subclass
- H04L69/16—Implementation or adaptation of Internet protocol [IP], of transmission control protocol [TCP] or of user datagram protocol [UDP]
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L69/00—Network arrangements, protocols or services independent of the application payload and not provided for in the other groups of this subclass
- H04L69/16—Implementation or adaptation of Internet protocol [IP], of transmission control protocol [TCP] or of user datagram protocol [UDP]
- H04L69/163—In-band adaptation of TCP data exchange; In-band control procedures
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Computer Security & Cryptography (AREA)
- Environmental & Geological Engineering (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The invention discloses a method and a system for identifying a packet loss type in data transmission of a reliable transmission protocol. The method comprises the following steps: a step 1, initializing a base round trip delay base-RTT and a base packet loss distance base-Dist; a step 2, obtaining a packet loss delay loss-RTT of packet loss in the data transmission and a packet loss distance loss-Dist of the packet loss to the former packet loss; and a step 3, judging the packet loss type according to a decision formula, wherein the decision formula comprises: when the loss-RTT is not smaller than K * based-RTT and the loss-Dist is not larger than the base-Dist, the packet loss is congestion packet loss; when the loss-RTT is smaller than K * based-RTT and the loss-Dist is larger than the base-Dist, the packet loss is random packet loss, and K is not smaller than 2 and is not larger than 5. The method disclosed by the invention can be used for accurately identifying the packet loss type of the packet loss to provide reference for the optimization of a TCP congestion control strategy. The method disclosed by the invention can also be applied to the single-end optimization of a TCP, which is convenient for the optimization and the performance improvement of the reliable transmission protocol.
Description
Technical field
The present invention relates to computer network field, particularly relate to a kind of distinguish the method and system of network packet packet loss type in reliable transport protocol.
Background technology
Common reliable transport protocol includes TCP, UDT etc. TCP (TransmissionControlProtocol transmission control protocol) is a kind of connection-oriented, reliable, transport layer communication protocol based on byte stream. UDT is the Data Transport Protocol (UDP-basedDataTransferProtocol of UDP, it is called for short UDT) it is a kind of based on the reliable data transmission agreement in the application layer realized on UDP basis, due to they reliability and towards characteristics such as connections, therefore apply in a network widely. HTTP, FTP, SMTP and TelNet etc. use Transmission Control Protocol, and in the network environment of high time delay bandwidth product, the application of UDT is also widely. The reliability of reliable transport protocol realizes mainly by packet sequence number, confirmation and retransmission mechanism, and the flow-control of agreement is mainly through windowing mechanism and what congestion control policy completed jointly.
Reliable transport protocol data stream in network transmission process owing to the restriction of link intermediate equipment disposal ability or the problem of link itself can produce packet loss phenomenon, generally the packet loss caused due to the restriction of intermediate equipment disposal ability or congestion control in link is referred to as congestion packet loss, and the packet loss phenomenon produced due to some non-congested control strategies such as link failure, packet check errors is referred to as random loss.
For a packet loss, present stage does not have effective means to be congestion packet loss or random loss to distinguish this packet loss, and packet loss is all identified by conventional jamming control method (Reno, Cubic etc. in TCP) as network congestion. So, it is directly entered congestion control to process if being left out its packet loss characteristic for a packet loss, utilization rate and the stability that can cause link decline serious, particularly in the chain environment that random loss rate is higher, not differentiating between packet loss type, seriously lowering link utilization thus affecting network performance (QualityofService).
For this situation, we have proposed one and identify packet loss type by mechanism, identified based on the type of packet loss in the data transmission of reliable transport protocol by some information datas in analytical data transmitting procedure. New congestion control policy can be designed on this basis, thus being easier to and effectively optimize the transmission performance of reliable transport protocol.
Summary of the invention
Present invention solves the technical problem that and be in that, identify based on the type of packet loss in the data transmission of reliable transport protocol.
Further, it is possible to the performance for promoting reliable transport protocol (TCP, UDT etc.) provides accurate reference information.
In order to solve the problems referred to above, the method that the invention discloses packet loss type in a kind of data identifying reliable transport protocol transmission, the method includes:
Step 1, initializes benchmark round-trip delay base-RTT and benchmark packet loss distance base-Dist;
Step 2, obtains the packet loss time delay loss-RTT of the packet loss occurred in the data transmission and the packet loss distance loss-Dist of this packet loss front packet loss apart;
Step 3, judges the type of this packet loss according to a decision-making formula, and this decision-making formula includes:
Work as loss-RTT >=K*base-RTT and loss-Dist��base-Dist, this packet loss is congestion packet loss;
As loss-RTT<K*base-RTT and loss-Dist>base-Dist, this packet loss is random loss;
Wherein, 2��K��5.
In this step 1,
base-RTT(i+1)=MIN (base-RTT(i),RTT(i+1))
Base-RTT (i+1) is the benchmark round-trip delay of i+1 data stream, and base-RTT (i) is the benchmark round-trip delay of i-th data stream, and RTT (i+1) is the round-trip delay of all packets in i+1 data stream.
This benchmark packet loss distance base-Dist is 2 or 3 or 4 or 5 or 6 or 7 or 8 or 9 or 10.
In this step 2, two packets continuously transmitted during data are transmitted, if the Pi packet fails to receive confirmation packet ACK, the Pi+1 packet can receive confirmation packet ACK, then the Pi packet is considered as packet loss, and its packet loss time delay loss-RTT is the round-trip delay of Pi+1 packet;
Or, if the Pi+1 packet fails to receive confirmation packet ACK, the Pi packet can receive confirmation packet ACK, then the Pi+1 packet is considered as packet loss, and its packet loss time delay loss-RTT is the round-trip delay of Pi packet.
This packet loss distance is that between two adjacent packet losses, the number of packet adds one. K=3.
The invention also discloses the system of packet loss type in a kind of data identifying reliable transport protocol transmission, this system includes:
Initialization module, for initializing benchmark round-trip delay base-RTT and benchmark packet loss distance base-Dist;
Parameter acquisition module, the packet loss distance loss-Dist of packet loss time delay loss-RTT and this packet loss front packet loss apart for obtaining the packet loss occurred in the data transmission;
Judge module, for judging the type of this packet loss according to a decision-making formula, this decision-making formula includes:
Work as loss-RTT >=K*base-RTT and loss-Dist��base-Dist, this packet loss is congestion packet loss;
As loss-RTT<K*base-RTT and loss-Dist>base-Dist, this packet loss is random loss;
Wherein, 2��K��5.
In this initialization module,
base-RTT(i+1)=MIN (base-RTT(i),RTT(i+1))
Base-RTT (i+1) is the benchmark round-trip delay of i+1 data stream, and base-RTT (i) is the benchmark round-trip delay of i-th data stream, and RTT (i+1) is the round-trip delay of all packets in i+1 data stream.
This benchmark packet loss distance base-Dist is 2 or 3 or 4 or 5 or 6 or 7 or 8 or 9 or 10.
Two packets that this parameter acquisition module continuously transmits in transmitting for data, if the Pi packet fails to receive confirmation packet ACK, the Pi+1 packet can receive confirmation packet ACK, then the Pi packet being considered as packet loss, its packet loss time delay loss-RTT is the round-trip delay of Pi+1 packet;
Or, if the Pi+1 packet fails to receive confirmation packet ACK, the Pi packet can receive confirmation packet ACK, then the Pi+1 packet is considered as packet loss, and its packet loss time delay loss-RTT is the round-trip delay of Pi packet.
This packet loss distance is that between two adjacent packet losses, the number of packet adds one. K=3.
The present invention can identify the packet loss type of packet loss accurately, it is judged that it belongs to congestion packet loss or random loss. This packet loss categorical data can provide foundation for the optimization of TCP congestion control policy. And control module and the congestion situation in network can be dredged based on this TCP congestion control policy, improve the efficiency of network transmission. It addition, the single-ended optimization of reliable transport protocol realizes only by adjusting congestion control policy, the method identifying packet loss type of the present invention can provide accurate reference information for congestion control policy, it is simple to the determination of congestion control policy. Therefore, the method identifying packet loss type of the present invention is also applied in the single-ended optimization of Transmission Control Protocol, it is simple to the optimization of reliable transport protocol and performance boost.
Accompanying drawing explanation
Fig. 1 show the schematic network structure of the present invention.
The schematic flow sheet of the method for the packet loss type that Fig. 2 show in a kind of data transmission identifying reliable transport protocol of the present invention.
Fig. 3 show benchmark RTT distribution situation schematic diagram.
Fig. 4 show data stream packet loss distribution situation schematic diagram.
Fig. 5 show data stream throughput distribution situation schematic diagram.
Fig. 6 show round-trip delay distribution situation schematic diagram during packet loss.
Fig. 7 show and assembles the curve chart of ratio shared by packet loss.
Fig. 8 show the distribution situation schematic diagram of continual data package dropout.
Detailed description of the invention
Below in conjunction with specific embodiment be described in detail the present invention realize process.
The present invention utilizes the packet loss time delay in the data transmission procedure of reliable transport protocol, packet loss distance, and distinguishing the packet loss occurred in reliable transport protocol data stream in real time is congestion packet loss or random loss.
It is illustrated in figure 1 the schematic network structure of the present invention. The schematic flow sheet of the method for the packet loss type being illustrated in figure 2 in a kind of data transmission identifying reliable transport protocol of the present invention.
Being carried out data transmission by reliable transport protocol between transmitting terminal and the receiving terminal of the present invention, data transmission is transmitted management by the control module embedded by transmitting terminal, by performing congestion control policy, promotes transmission performance.
The method of the packet loss type in the data transmission of this identification reliable transport protocol includes:
Step 1, initializes benchmark round-trip delay base-RTT and benchmark packet loss distance base-Dist.
Specifically, RTT represents from transmitting terminal transmission data, receives the confirmation packet (ACK) from receiving terminal, the time delay altogether experienced to transmitting terminal.
Benchmark round-trip delay base-RTT represents the minimum RTT that can reach in tcp data transmitting procedure. For data cube computation or the reliable transport protocol data stream of each reliable transport protocol, all set a base-RTT.
RTT size and for the value of each base-RTT, its initial value is between transmitting terminal and receiving terminal, in setting up connection procedure. For the connection procedure based on TCP, this base-RTT is RTT size during three-way handshake.
And in data transmission procedure, base-RTT dynamically changes according to the RTT situation of change of packet in data stream, base-RTT takes the minima of the RTT of current stream in data transmission procedure.
In one embodiment:
base-RTT(i+1)=MIN (base-RTT(i),RTT(i+1))(1)
base-RTT(i+1)For the benchmark round-trip delay of i+1 data stream, base-RTT(i)For the benchmark round-trip delay of i-th data stream, RTT(i+1)For the round-trip delay of all packets in i+1 data stream.
For benchmark packet loss distance base-Dist, looking first at packet distance: naturally arranged according to transmission order by the packet in all of transmitting procedure, the number of the packet between two packets adds one and is two packet distances. Such as the packet distance between the 2nd packet and the 5th packet is 3. And the number of the packet that packet loss distance is between two adjacent packet losses adds one. The value of this benchmark packet loss distance base-Dist can be 2 or 3 or 4 or 5 or 6 or 7 or 8 or 9 or 10.
Step 2, obtains the packet loss time delay loss-RTT of the packet loss occurred in the data transmission and the packet loss distance loss-Dist of this packet loss front packet loss apart.
In the process that transmitting terminal carries out data transmission based on reliable transport protocol, it is possible to obtain current data wraps in the transmission parameter in transmitting procedure, then can obtain the information of the packet of normal transmission from this transmission parameter, and then obtain packet loss information. For the packet normally sent, it is possible to know its round-trip delay.
This transmission parameter can be sent to this control module by transmitting terminal, this control module extract packet loss time delay loss-RTT and packet loss distance loss-Dist from this transmission parameter and carry out subsequent calculations.
Specifically, two packet P of any pair continuous print in the data stream of a reliable transport protocoli��Pi+1, there is following four and be likely to situation:
1, two packets are all properly received and return corresponding ACK.
2��PiLose, Pi+1It is properly received and returns corresponding ACK.
3��Pi+1Lose, PiIt is properly received and returns corresponding ACK.
4��PiLose and Pi+1Losing, this kind of situation is continual data package dropout phenomenon.
For situation 1, it is absent from packet loss, it is not necessary to distinguish packet loss type.
For situation 2, packet PiPacket loss, now, defines packet PiPacket loss time delay loss-RTT be RTT (Pi+1), namely with packet Pi+1The value of round-trip delay as packet loss PiPacket loss time delay loss-RTT.
For situation 3, packet Pi+1Packet loss, now, defines packet Pi+1Packet loss time delay loss-RTT be RTT (Pi), namely packet PiThe value of round-trip delay as packet loss Pi+1Packet loss time delay loss-RTT.
For situation 4, follow-up it is analyzed again.
Packet loss time delay loss-RTT the rule of utilization power 2 or the rule of situation 3 can be carried out value by the present invention.
Owing to can know which packet there occurs packet loss based on reliable transport protocol, so the packet loss distance loss-Dist obtaining each packet loss apart from a front packet loss can be calculated.
Step 3, this control module judges the type of this packet loss according to a decision-making formula, and this decision-making formula includes:
Work as loss-RTT >=K*base-RTT and loss-Dist��base-Dist, this packet loss is congestion packet loss;
As loss-RTT<K*base-RTT and loss-Dist>base-Dist, this packet loss is random loss;
Wherein, 2��K��5. Further, K is preferably 3.
Wherein, if loss-Dist��base-Dist, current packet loss is defined for assembling packet loss (Agg-Loss); If it is scattered packet loss (Odd-Loss) that loss-Dist > base-Dist defines current packet loss.
Hereinafter the formula in above-mentioned steps 3 is arranged and make an explanation.
Inventor analyzes the package level data flow data of certain cloud storage service provider, by the calculating to all TCP flow, the distribution situation of its benchmark RTT is as shown in Figure 3, can draw from this figure, majority of traffic (>=70%) the benchmark round-trip delay of data stream all between 10ms and 50ms, and only the benchmark round-trip delay of data stream of about 10% is more than 100ms. The RTT size illustrated in Ethernet in the data transmission procedure of reliable transport protocol is not more than 100ms.
On the other hand, when the packet drop of all of tcp data stream is analyzed, as shown in Figure 4, the data stream of about 50% is absent from packet loss phenomenon in transmitting procedure, and the data stream of about 15%, in its data transmission procedure, packet loss (lossrate) is less than 2.5%, and in data transmission procedure, the packet loss data stream more than 10% accounts for 10%. Owing to different data stream is through no transmission path, or the congestion packet loss of network intermediary device has tendentiousness, so creating packet loss different in link, thus causing the difference of the transmission performance of various flows.
Fig. 5 show data stream throughput distribution situation schematic diagram. The performance of every data stream throughput described in figure. Can be A line (packet loss) and B line (without packet loss) by data flow point according to whether there is packet loss, in the drawings, on the whole in all of packet data stream throughput more than 90% more than 30KB/s. The throughput data stream more than 100KB/s is being about 50% by the ratio in packet loss data stream, and is being about 70% without the ratio in the data stream of packet loss. On the other hand without in packet loss data stream, there are about the throughput of the data stream of 10% and can reach 1MB/s, and in packet loss stream, the proportion of data stream more than 1MB/s is 0, can draw from this result, the packet loss in link can significantly reduce the transmission performance of data stream. This is because the congestion control policy in link is adjusted congestion window according to packet loss or link actual treatment ability causes hydraulic performance decline.
Further the packet loss of certain cloud storage service provider is analyzed, the ratio of packet loss round-trip delay during using lost data packets and benchmark round-trip delay is as x-axis, the shared ratio of accumulation is as y-axis, the ratio calculating all data obtains as shown in Figure 6, from this figure it can be seen that, the ratio of the packet loss loss-RTT/base-RTT of about more than 90% 2 and more than, the ratio of the packet loss loss-RTT/base-RTT of about more than 80% 3 and more than, the ratio of the packet loss loss-RTT/base-RTT of about 50% is 5. Owing to the quantity of random loss is less, then the data in Fig. 6 can be accordingly to be regarded as and be produced by congestion packet loss. From fig. 6 it can be seen that network congestion not only creates substantial amounts of packet loss and adds the packet time in intermediate equipment relief area, thus when causing compared with benchmark round-trip delay several times round.
Owing to network congestion is easily generated the RTT that multiple increases, then the multiple of packet loss RTT relative datum RTT can make to estimate a key property of a reliable transport protocol packet packet loss.
Therefore, above-mentioned K may be selected to be 2��K��5. In this interval, cover the packet loss of more than 50%, it is possible to as the basis differentiating congestion packet loss.
Additionally, investigate by all of tcp data stream packet loss in above-mentioned data is assembled situation, loss-Dist is respectively set to 1,3,5, can obtain assembling the curve chart of ratio shared by packet loss, as shown in Figure 7, it will be seen that in the figure, the proportional curve of loss-Dist=3 and loss-Dist=5 gained overlaps.
Axis of abscissas in Fig. 7 represents the ratio assembling the shared total packet loss of packet loss in stream, and what ordinate axle represented is meet the ratio that the stream of axis of abscissas ratio occupies at all of data stream.
As we know from the figure, when loss-Dist is 1, the data stream (vertical coordinate 0.5) of about 50%, it assembles ratio shared by packet loss more than 75% (abscissa 0.75), and when loss-Dist is 3 and 5, assembling the ratio of the ratio stream (abscissa 0.75) more than 75% shared by packet loss is 70%, therefore for congestion packet loss, owing to link processing ability fault is easier to occur to assemble packet loss. So RTT change and packet loss characteristic during in conjunction with packet loss distinguish random loss and congestion packet loss easily.
For the concrete value of base-Dist, first it have to be larger than 1, if equal to 1, inevitable corresponding continual data package dropout, then there is no too big differentiation meaning. Thus can from 2 integer values started. If arrange too small, it is easy to make part congestion packet loss be identified as random loss, and if arrange excessive, the part packet loss in random loss can be made to be identified as congestion packet loss.
When base-Dist is set to 3, it is identical for assembling distribution proportion in packet loss proportion and a little scope more than 3. Such as, in Fig. 7, the proportional curve of loss-Dist=3 and loss-Dist=5 gained overlaps. Thus, value is 3 or 5, it is possible to correspondence identifies most congestion packet loss, can as preferably.
Base-Dist can also be the integer less than or equal to 10.
It addition, control module also can according to network demand, to loss-RTT>=K*base-RTT and loss-Dist>base-Dist, or packet loss type during loss-RTT<K*base-RTT and loss-Dist��base-Dist is determined. This situation is actually considerably less, because congestion packet loss two conditions are usually and change simultaneously.
If network demand is more radical, it is possible to above-mentioned situation is classified as random loss, if network demand is more conservative, it is possible to above-mentioned situation is classified as congestion packet loss. Fixed as desired.
Continual data package dropout for situation 4, by to tcp data stream bag data analysis substantial amounts of in above-mentioned data, obtaining as shown in Figure 8, the packet of about 56% broadly falls into continual data package dropout (continual data package dropout length is be more than or equal to 2), there are about the continual data package dropout length of 20% be 4 and more than. And when continual data package dropout length is bigger, sender can not receive ACK (or SACK) for a long time, thus causing that the time out timer of packet surpasses, the packet of described continuous loss now by Retransmission timeout, will be recovered. Packet owing to losing obtains recovery, thus is absent from packet loss, without the judgement carrying out packet loss type. Visible, situation 4 is identical with situation 1, all without carrying out the judgement of packet loss type.
By said method, the present invention can identify the packet loss type of packet loss accurately, it is judged that it belongs to congestion packet loss or random loss. Foundation can be provided to the optimization of TCP congestion control policy based on this packet loss categorical data. And control module and the congestion situation in network can be dredged based on this TCP congestion control policy, improve the efficiency of network transmission.
It addition, the single-ended optimization of reliable transport protocol realizes only by adjusting congestion control policy, the method identifying packet loss type of the present invention can provide accurate reference information for congestion control policy, it is simple to the determination of congestion control policy. Therefore, the method identifying packet loss type of the present invention is also applied in the single-ended optimization of Transmission Control Protocol, it is simple to the optimization of reliable transport protocol and performance boost.
The foregoing is only the preferred embodiments of the present invention, be not intended to the scope of patent protection of the limitation present invention, therefore such as use the equivalence that present invention is done to change, be all in like manner all included within the scope of the present invention.
Claims (12)
1. the method for the packet loss type identified in the data transmission of reliable transport protocol, it is characterised in that the method includes:
Step 1, initializes benchmark round-trip delay base-RTT and benchmark packet loss distance base-Dist;
Step 2, obtains the packet loss time delay loss-RTT of the packet loss occurred in the data transmission and the packet loss distance loss-Dist of this packet loss front packet loss apart;
Step 3, judges the type of this packet loss according to a decision-making formula, and this decision-making formula includes:
Work as loss-RTT >=K*base-RTT and loss-Dist��base-Dist, this packet loss is congestion packet loss;
As loss-RTT<K*base-RTT and loss-Dist>base-Dist, this packet loss is random loss;
Wherein, 2��K��5.
2. the method for claim 1, it is characterised in that in this step 1,
base-RTT(i+1)=MIN (base-RTT(i),RTT(i+1))
base-RTT(i+1)For the benchmark round-trip delay of i+1 data stream, base-RTT(i)For the benchmark round-trip delay of i-th data stream, RTT(i+1)For the round-trip delay of all packets in i+1 data stream.
3. the method for claim 1, it is characterised in that this benchmark packet loss distance base-Dist is 2 or 3 or 4 or 5 or 6 or 7 or 8 or 9 or 10.
4. the method for claim 1, it is characterised in that in this step 2, two packets continuously transmitted during data are transmitted, if PiIndividual packet fails to receive confirmation packet ACK, Pi+1Individual packet can receive confirmation packet ACK, then PiIndividual packet is considered as packet loss, and its packet loss time delay loss-RTT is Pi+1The round-trip delay of individual packet;
Or, if Pi+1Individual packet fails to receive confirmation packet ACK, PiIndividual packet can receive confirmation packet ACK, then Pi+1Individual packet is considered as packet loss, and its packet loss time delay loss-RTT is PiThe round-trip delay of individual packet.
5. the method for claim 1, it is characterised in that this packet loss distance is that between two adjacent packet losses, the number of packet adds.
6. the method for claim 1, it is characterised in that K=3.
7. the system of the packet loss type identified in the data transmission of reliable transport protocol, it is characterised in that this system includes:
Initialization module, for initializing benchmark round-trip delay base-RTT and benchmark packet loss distance base-Dist;
Parameter acquisition module, the packet loss distance loss-Dist of packet loss time delay loss-RTT and this packet loss front packet loss apart for obtaining the packet loss occurred in the data transmission;
Judge module, for judging the type of this packet loss according to a decision-making formula, this decision-making formula includes:
Work as loss-RTT >=K*base-RTT and loss-Dist��base-Dist, this packet loss is congestion packet loss;
As loss-RTT<K*base-RTT and loss-Dist>base-Dist, this packet loss is random loss;
Wherein, 2��K��5.
8. system as claimed in claim 7, it is characterised in that in this initialization module,
base-RTT(i+1)=MIN (base-RTT(i),RTT(i+1))
base-RTT(i+1)For the benchmark round-trip delay of i+1 data stream, base-RTT(i)For the benchmark round-trip delay of i-th data stream, RTT(i+1)For the round-trip delay of all packets in i+1 data stream.
9. system as claimed in claim 7, it is characterised in that this benchmark packet loss distance base-Dist is 2 or 3 or 4 or 5 or 6 or 7 or 8 or 9 or 10.
10. system as claimed in claim 7, it is characterised in that two packets that this parameter acquisition module continuously transmits in transmitting for data, if PiIndividual packet fails to receive confirmation packet ACK, Pi+1Individual packet can receive confirmation packet ACK, then by PiIndividual packet is considered as packet loss, and its packet loss time delay loss-RTT is Pi+1The round-trip delay of individual packet;
Or, if Pi+1Individual packet fails to receive confirmation packet ACK, PiIndividual packet can receive confirmation packet ACK, then by Pi+1Individual packet is considered as packet loss, and its packet loss time delay loss-RTT is PiThe round-trip delay of individual packet.
11. system as claimed in claim 7, it is characterised in that this packet loss distance is that between two adjacent packet losses, the number of packet adds.
12. system as claimed in claim 7, it is characterised in that K=3.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610082255.6A CN105634875A (en) | 2016-02-05 | 2016-02-05 | Method and system for identifying packet loss type in data transmission of reliable transmission protocol |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610082255.6A CN105634875A (en) | 2016-02-05 | 2016-02-05 | Method and system for identifying packet loss type in data transmission of reliable transmission protocol |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105634875A true CN105634875A (en) | 2016-06-01 |
Family
ID=56049405
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610082255.6A Pending CN105634875A (en) | 2016-02-05 | 2016-02-05 | Method and system for identifying packet loss type in data transmission of reliable transmission protocol |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105634875A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2017132987A1 (en) * | 2016-02-05 | 2017-08-10 | 中国科学院计算技术研究所 | Method and system for recognizing packet loss type in data transmission of reliable transmission protocol |
| CN107634908A (en) * | 2016-07-19 | 2018-01-26 | 华为技术有限公司 | A kind of method and apparatus of data transfer |
| CN109104742A (en) * | 2017-06-20 | 2018-12-28 | 华为技术有限公司 | Congestion window method of adjustment and sending device |
| CN109474538A (en) * | 2018-12-29 | 2019-03-15 | 北京达佳互联信息技术有限公司 | A kind of data transmission method, device, terminal device and storage medium |
| CN113595830A (en) * | 2021-07-30 | 2021-11-02 | 百果园技术(新加坡)有限公司 | Method, device, equipment and storage medium for detecting network packet loss state |
| CN114650260A (en) * | 2022-03-14 | 2022-06-21 | 清华大学 | Network packet loss type identification method and device and electronic equipment |
| CN115086200A (en) * | 2022-05-06 | 2022-09-20 | 北京达佳互联信息技术有限公司 | Packet loss type determination method and device, electronic equipment and storage medium |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101115002A (en) * | 2007-03-19 | 2008-01-30 | 重庆邮电大学 | Method for improving wireless self-organizing network performance using TCP Veno |
| EP2256991A1 (en) * | 2009-05-25 | 2010-12-01 | Canon Kabushiki Kaisha | Method and device for determining types of packet loss in a communication network |
| CN104113884A (en) * | 2013-04-18 | 2014-10-22 | 南京邮电大学 | Real-time multimedia transmission rate control mechanism in wireless network |
| CN104486690A (en) * | 2014-12-25 | 2015-04-01 | 北京理工大学 | Mobile video transmission optimization method based on TCP (transmission control protocol) |
-
2016
- 2016-02-05 CN CN201610082255.6A patent/CN105634875A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101115002A (en) * | 2007-03-19 | 2008-01-30 | 重庆邮电大学 | Method for improving wireless self-organizing network performance using TCP Veno |
| EP2256991A1 (en) * | 2009-05-25 | 2010-12-01 | Canon Kabushiki Kaisha | Method and device for determining types of packet loss in a communication network |
| CN104113884A (en) * | 2013-04-18 | 2014-10-22 | 南京邮电大学 | Real-time multimedia transmission rate control mechanism in wireless network |
| CN104486690A (en) * | 2014-12-25 | 2015-04-01 | 北京理工大学 | Mobile video transmission optimization method based on TCP (transmission control protocol) |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2017132987A1 (en) * | 2016-02-05 | 2017-08-10 | 中国科学院计算技术研究所 | Method and system for recognizing packet loss type in data transmission of reliable transmission protocol |
| CN107634908A (en) * | 2016-07-19 | 2018-01-26 | 华为技术有限公司 | A kind of method and apparatus of data transfer |
| CN109104742A (en) * | 2017-06-20 | 2018-12-28 | 华为技术有限公司 | Congestion window method of adjustment and sending device |
| CN109474538A (en) * | 2018-12-29 | 2019-03-15 | 北京达佳互联信息技术有限公司 | A kind of data transmission method, device, terminal device and storage medium |
| CN109474538B (en) * | 2018-12-29 | 2021-07-30 | 北京达佳互联信息技术有限公司 | Data transmission method and device, terminal equipment and storage medium |
| CN113595830A (en) * | 2021-07-30 | 2021-11-02 | 百果园技术(新加坡)有限公司 | Method, device, equipment and storage medium for detecting network packet loss state |
| CN113595830B (en) * | 2021-07-30 | 2024-02-20 | 百果园技术(新加坡)有限公司 | Method, device, equipment and storage medium for detecting network packet loss state |
| CN114650260A (en) * | 2022-03-14 | 2022-06-21 | 清华大学 | Network packet loss type identification method and device and electronic equipment |
| CN114650260B (en) * | 2022-03-14 | 2024-03-12 | 清华大学 | Network packet loss type identification method and device and electronic equipment |
| CN115086200A (en) * | 2022-05-06 | 2022-09-20 | 北京达佳互联信息技术有限公司 | Packet loss type determination method and device, electronic equipment and storage medium |
| CN115086200B (en) * | 2022-05-06 | 2023-09-01 | 北京达佳互联信息技术有限公司 | Packet loss type determining method and device, electronic equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105634875A (en) | Method and system for identifying packet loss type in data transmission of reliable transmission protocol | |
| CN102098301B (en) | The adaptive data transmission method of multilink and system | |
| CN106027412B (en) | A kind of TCP jamming control method based on congestion queue length | |
| CN103999409B (en) | Method for link buffer size and queue length estimation for bandwidth-varying mobile data networks | |
| CN101645765B (en) | Reliable transmission acceleration method facing networks with high error rate and long delay characteristics | |
| CN103457871B (en) | Based on the increasing window method of the congestion avoidance phase of deferred constraint in DCN | |
| US20020078164A1 (en) | System and method for data transfer acceleration in a TCP network environment | |
| CN102595509B (en) | Cocurrent data distribution method based on transmission control protocol (TCP) in heterogeneous networks | |
| CN106982108A (en) | The method and relevant device of a kind of data transfer | |
| CN105049369A (en) | Video transmission congestion control method based on MPTCP in heterogeneous wireless network | |
| CN105024940A (en) | Link adaptation-based heterogeneous network TCP congestion control method | |
| CN101964755A (en) | Explicit congestion control method based on bandwidth estimation in high-bandwidth delay network | |
| CN106059951A (en) | Transmission control method for DCN (Data Center Network) based on multilevel congestion feedback | |
| Wang et al. | TCP congestion control algorithm for heterogeneous Internet | |
| CN103986548A (en) | Method for determining packet loss reasons and terminal | |
| CN105933242B (en) | A kind of method and system improving data center TCP response speed | |
| CN103763212A (en) | Method for regulating network congestion | |
| CN103281287A (en) | User datagram protocol (UDP)-based wind generating set communication method and system | |
| WO2017132987A1 (en) | Method and system for recognizing packet loss type in data transmission of reliable transmission protocol | |
| Gu et al. | Udt: An application level transport protocol for grid computing | |
| Zhou et al. | TCP stalls at the server side: Measurement and mitigation | |
| CN107147585A (en) | A kind of flow control methods and device | |
| CN103595552A (en) | Analysis method and analysis system for cluster storage network parallel load | |
| Zou et al. | Gtcp: Hybrid congestion control for cross-datacenter networks | |
| Meng et al. | Research on TCPW improvement in hybrid network |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20160601 |
|
| RJ01 | Rejection of invention patent application after publication |