CN108540323B - Method for predicting processing rate of router based on minimum additive deconvolution - Google Patents

Method for predicting processing rate of router based on minimum additive deconvolution Download PDF

Info

Publication number
CN108540323B
CN108540323B CN201810315283.7A CN201810315283A CN108540323B CN 108540323 B CN108540323 B CN 108540323B CN 201810315283 A CN201810315283 A CN 201810315283A CN 108540323 B CN108540323 B CN 108540323B
Authority
CN
China
Prior art keywords
flow
data packets
node
data packet
router node
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201810315283.7A
Other languages
Chinese (zh)
Other versions
CN108540323A (en
Inventor
易运晖
杨新艳
赵楠
陈南
孟艳红
朱畅华
权东晓
韩宝彬
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Xidian University
Original Assignee
Xidian University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Xidian University filed Critical Xidian University
Publication of CN108540323A publication Critical patent/CN108540323A/en
Application granted granted Critical
Publication of CN108540323B publication Critical patent/CN108540323B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/08—Configuration management of networks or network elements
    • H04L41/0803—Configuration setting
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/14—Network analysis or design
    • H04L41/142—Network analysis or design using statistical or mathematical methods
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/14—Network analysis or design
    • H04L41/147—Network analysis or design for predicting network behaviour
    • 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
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/01—Protocols
    • H04L67/06—Protocols specially adapted for file transfer, e.g. file transfer protocol [FTP]

Landscapes

  • Engineering & Computer Science (AREA)
  • Signal Processing (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • General Physics & Mathematics (AREA)
  • Physics & Mathematics (AREA)
  • Algebra (AREA)
  • Environmental & Geological Engineering (AREA)
  • Mathematical Analysis (AREA)
  • Mathematical Optimization (AREA)
  • Mathematical Physics (AREA)
  • Probability & Statistics with Applications (AREA)
  • Pure & Applied Mathematics (AREA)
  • Data Exchanges In Wide-Area Networks (AREA)

Abstract

The invention discloses a method for predicting the processing rate of a router based on minimum additive deconvolution, which comprises the following steps: (1) building a computer network; (2) configuring data packet transmission of a computer network; (3) collecting statistics; (4) counting the arrival flow and the departure flow of the router; (5) calculating the total number of arriving data packets in unit time and the total flow of the data packets; (6) the method comprises the steps of (1) calculating the rate of the router processing the data packet flow at each moment in the (0, s) period by utilizing minimum addition deconvolution, and (7) predicting the processing rate of the router to the computer network data packet flow.

Description

Method for predicting processing rate of router based on minimum additive deconvolution
Technical Field
The invention belongs to the technical field of communication, and further relates to a method for predicting the processing rate of a router based on minimum additive deconvolution in the technical field of computer networks. The invention predicts the processing rate of the router node by utilizing network calculus and minimum deconvolution according to the arrival flow and the departure flow of the router node in the computer network, and can be used for solving the problem of processing rate prediction of the router node in the computer network.
Background
The network computation is a deterministic queuing theory based on nonlinear algebra, can calculate the bounds of performance parameters such as delay, backlog and the like after the traffic of a data packet sent by a source node reaches a destination node through a router node, and can carry out quantitative analysis on the traffic problem in the network. Different scheduling policies, priority permissions, etc. of the router nodes will cause the router nodes to process different rates for each data stream. In the prior art, the maximum processing rate of a router node is mostly known, the minimum boundary and the maximum boundary of leaving traffic are deduced based on the arrival traffic and the maximum processing rate which can be provided by the router node, and then backlog and delay are calculated by using minimum addition operation.
The university of science and technology in china discloses a method for predicting the end-to-end delay of the fast forwarding service in the patent document "a method for predicting the end-to-end delay of the fast forwarding service" (patent application No. 200910060549.9, publication No. CN 101478456). The method sets data flow as poisson flow, firstly shapes the data flow before entering a network to make the data flow obey the same arrival curve, obtains the arrival curve of the fast forwarding data flow by using a deterministic network operation theory, does not consider the processing rate of a router node, then calculates the delay upper bound of a single node, and finally obtains the probability distribution expression of the end-to-end delay upper bound of the fast forwarding data flow. The method has the disadvantages that as the invention requires the data flow to be shaped before entering the network so as to be obeyed to the same arrival curve, for the router node, a plurality of different data flows generally arrive, and different processing rates are provided for the different data flows, so that the problems of processing rate prediction and time delay backlog calculation when the plurality of data flows arrive can not be effectively solved.
The patent document applied by Beijing aerospace university "AFDX end-delay upper bound calculation method based on random network calculus" (patent application No. 201210253774.6, publication No. CN102780581A) discloses an AFDX end-delay upper bound calculation method based on random network calculus. The scheduling strategy used by the router node in the method is non-preemptive scheduling, a fixed bandwidth is set, namely the router node processes the data packet flow at a fixed speed, then the arrival curve of the data flow is determined, and finally the end time delay is analyzed and predicted. The method has the defects that the router node provides the maximum processing rate when the data flow is large enough, and the calculation of the time delay by the maximum processing rate when the data flow is small has larger errors, so that the method can not be used for accurately solving the problems of the processing rate and the time delay test of the router node in the common computer network.
Disclosure of Invention
The invention aims to provide a method for predicting the processing rate of a router based on minimum additive deconvolution, which can predict the processing rate of a router node on the network traffic of a computer by calculating the rate of the router node processing the data packet traffic at each moment in a (0, s) period by using the minimum additive deconvolution, and solves the problem of predicting the processing rate of the router node in the network under the condition that the processing rate of the router node is unknown.
In order to achieve the purpose, the method comprises the following specific steps:
(1) building a computer network:
connecting two user nodes and a router node by using a link to build a computer network;
(2) configuring data packet transmission of a computer network:
taking any one user node as a source node, taking another user node as a destination node, leading the source node to obey the exponential distribution according to the packet interval, generating a plurality of file transfer FTP data packets according to the rule of 5000KB according to the packet size, and forwarding the file transfer FTP data packets to the destination node through the router node;
(3) collecting statistics:
a source node and a destination node, which respectively collect the data packet flow sent and received at each time within the (0, s) period, wherein s represents the time when the router node finishes processing the last data packet;
(4) and (3) counting the arrival traffic and the departure traffic of the router nodes:
(4a) counting the flow of a data packet sent by a source node reaching a router node;
(4b) after the data packet flow reaches the router node, the router node sequentially processes the data packets according to first-in first-out, and counts the data packet flow leaving the router node;
(5) calculating the total number of data packets arriving in unit time and the total flow of the data packets:
(5a) solving the total number of data packets arriving at the router node and the total flow of the arriving data packets by a least square curve fitting method, wherein the cumulative fitting flow of the arriving data packets of the router node reaches the router node in unit time;
(5b) solving the total number of data packets which leave the router node and reach the destination node in unit time of cumulative fitting flow of the data packets and the total flow of the arrived data packets by using a least square curve fitting method;
(6) the rate at which the router node processes packet traffic at each time during the (0, s) period is calculated using the following minimum additive deconvolution equation:
Figure GDA0002303436960000031
wherein β (t) represents the rate of the router node processing the data packet traffic at time t, 0 < t ≦ s, s represents the time when the router node has finished processing the last data packet, D (t) represents the cumulative fitting traffic of the router node leaving the data packet at time t, A (t) represents the cumulative fitting traffic of the router node arriving the data packet at time t,
Figure GDA0002303436960000032
representing a minimum add deconvolution operation;
(7) predicting the processing rate of the router node on the computer network data packet traffic:
and taking the maximum rate in all rates of the router node processing the data packet traffic in the (0, s) period as the processing rate of the router node on the computer network data packet traffic.
Compared with the prior art, the invention has the following advantages:
firstly, the method directly fits the arriving data packet flow and the leaving data packet flow of the router node by using a least square fitting curve method, then calculates the rate of the router node processing the data packet flow at each moment in the (0, s) period by using minimum additive deconvolution, predicts the processing rate of the router node on the computer network data packet flow, and overcomes the problem that the data flow must be shaped before entering the network to be subject to the same arriving curve in the prior art, so that the method has the advantages of simple operation, less constraint conditions and accurate prediction on the processing rate of the router node.
Secondly, the invention adopts a minimum deconvolution method, and utilizes the arriving data packet flow and the leaving data packet flow of the router node to calculate and predict the processing rate of the router node, thereby overcoming the problems that the data packet flow is processed at the maximum processing rate of the router node under the condition that the processing rate of the router node is unknown, and the prediction error is larger, so that the invention has the advantages of simple calculation process and small error.
Drawings
FIG. 1 is a flow chart of the present invention;
fig. 2 is a schematic diagram of a computer network constructed by the present invention.
Detailed Description
The invention is further described below with reference to the accompanying drawings. A
The specific steps implemented by the present invention are further described with reference to the flow chart of the present invention shown in fig. 1.
Step 1, building a computer network.
Two user nodes and a router node are connected by links to build a computer network as shown in figure 2.
And 2, configuring data packet transmission of the computer network.
Any user node is used as a source node, the other user node is used as a destination node, the source node obeys exponential distribution according to packet intervals, the size of the packets generates a plurality of file transfer FTP data packets according to the rule of 5000KB, and the file transfer FTP data packets are forwarded to the destination node through the router node.
And 3, collecting statistics.
The source node and the destination node respectively collect the data packet flow respectively transmitted and received at each moment in the (0, s) period, and s represents the moment when the router node finishes processing the last data packet.
And 4, counting the arrival flow and the departure flow of the router nodes.
And counting the flow of the data packet sent by the source node reaching the router node.
And after the data packet flow reaches the router node, the router node sequentially processes the data packets according to first-in first-out, and counts the data packet flow leaving the router node.
And 5, calculating the total number of the data packets arriving in the unit time and the total flow of the data packets.
And solving the total number of the data packets arriving at the router node and the total flow of the arriving data packets by using a least square curve fitting method, wherein the cumulative fitting flow of the arriving data packets of the router node reaches the router node in unit time.
And solving the total number of the data packets which leave the router node and reach the destination node in unit time according to the cumulative fitting flow of the data packets and the total flow of the arrived data packets by using a least square curve fitting method.
The least square curve fitting method comprises the following specific steps:
the method comprises the steps that firstly, data packet accumulated flow of configured source nodes and destination nodes in a built computer network in a sending time (0, s) period is collected;
subtracting the data packet cumulative fitting flow value corresponding to each moment from the data packet cumulative flow value at each moment, and taking the difference value as an error value at each moment;
thirdly, performing superposition operation after the error value of each moment is squared, and taking the operation result as the sum of the squared errors;
fourthly, solving the error square sum and the partial derivative to obtain the total number of the data packets of which the cumulative fitting flow of the data packets arrives in unit time;
and fifthly, obtaining the sum of squares of errors and partial derivatives to obtain the total number of the data packets of which the cumulative fitting flow of the data packets arrives in the unit time, and multiplying the total number of the data packets by the size of the data packets to obtain the total flow of the data packets of which the cumulative fitting flow of the data packets arrives in the unit time.
And 6, calculating the rate of the router node processing the data packet flow at each moment in the (0, s) period by using the following minimum addition deconvolution formula.
Wherein β (t) represents the rate of the router node processing the data packet traffic at time t, 0 < t ≦ s, s represents the time when the router node has finished processing the last data packet, D (t) represents the cumulative fitting traffic of the router node leaving the data packet at time t, A (t) represents the cumulative fitting traffic of the router node arriving the data packet at time t,
Figure GDA0002303436960000052
representing a minimum add deconvolution operation.
The arrived data packet accumulation fitting flow A (t) of the router node is used as an arrival curve of the router node, the departed data packet accumulation fitting flow D (t) of the router node is used as a departed curve of the router node, and β (t) is used as a service curve of the router node.
According to the output theorem of network calculus, the method can obtain
Figure GDA0002303436960000053
Will be provided with
Figure GDA0002303436960000054
As the upper boundary of d (t).
Obtaining the upper boundary of D (t) according to the exchange law of minimum convolution
Will be equation
Figure GDA0002303436960000056
Both sides are deconvoluted with A (t) minimum to obtain
Figure GDA0002303436960000057
According to the theorem of minimum addition deconvolution to obtain
Figure GDA0002303436960000058
And 7, predicting the processing rate of the router node to the computer network data packet traffic.
And taking the maximum rate of the router node for processing the data packet traffic in the (0, s) period as the processing rate of the router node for the data packet traffic in the computer network.

Claims (1)

1. A method for predicting a processing rate of a router based on minimum additive deconvolution, comprising the steps of:
(1) building a computer network:
connecting two user nodes and a router node by using a link to build a computer network;
(2) configuring data packet transmission of a computer network:
taking any one user node as a source node, taking another user node as a destination node, leading the source node to obey the exponential distribution according to the packet interval, generating a plurality of file transfer FTP data packets according to the rule of 5000KB according to the packet size, and forwarding the file transfer FTP data packets to the destination node through the router node;
(3) collecting statistics:
a source node and a destination node, which respectively collect the data packet flow sent and received at each time within the (0, s) period, wherein s represents the time when the router node finishes processing the last data packet;
(4) and (3) counting the arrival traffic and the departure traffic of the router nodes:
(4a) counting the flow of a data packet sent by a source node reaching a router node;
(4b) after the data packet flow reaches the router node, the router node sequentially processes the data packets according to first-in first-out, and counts the data packet flow leaving the router node;
(5) calculating the total number of data packets arriving in unit time and the total flow of the data packets:
(5a) solving the total number of data packets arriving at the router node and the total flow of the arriving data packets by a least square curve fitting method, wherein the cumulative fitting flow of the arriving data packets of the router node reaches the router node in unit time;
(5b) solving the total number of data packets which leave the router node and reach the destination node in unit time of cumulative fitting flow of the data packets and the total flow of the arrived data packets by using a least square curve fitting method;
the least square curve fitting method in the step (5a) and the step (5b) comprises the following specific steps:
the method comprises the steps that firstly, data packet accumulated flow of configured source nodes and destination nodes in a built computer network in a sending time (0, s) period is collected;
subtracting the data packet cumulative fitting flow value corresponding to each moment from the data packet cumulative flow value at each moment, and taking the difference value as an error value at each moment;
thirdly, performing superposition operation after the error value of each moment is squared, and taking the operation result as the sum of the squared errors;
fourthly, solving the error square sum and the partial derivative to obtain the total number of the data packets of which the cumulative fitting flow of the data packets arrives in unit time;
fifthly, the sum of the squares of the errors is subjected to partial derivation to obtain the total number of the data packets of which the cumulative fitting flow of the data packets arrives in unit time, and the total flow of the data packets of which the cumulative fitting flow of the data packets arrives in unit time is obtained by multiplying the size of the data packets by the total flow of the data packets;
(6) the rate at which the router node processes packet traffic at each time during the (0, s) period is calculated using the following minimum additive deconvolution equation:
Figure FDA0002303436950000021
wherein β (t) represents the rate of the router node processing the data packet traffic at time t, 0 < t ≦ s, s represents the time when the router node has finished processing the last data packet, D (t) represents the cumulative fitting traffic of the router node leaving the data packet at time t, A (t) represents the cumulative fitting traffic of the router node arriving the data packet at time t,
Figure FDA0002303436950000022
representing a minimum add deconvolution operation;
(7) predicting the processing rate of the router node on the computer network data packet traffic:
and taking the maximum rate in all rates of the router node processing the data packet traffic in the (0, s) period as the processing rate of the router node on the computer network data packet traffic.
CN201810315283.7A 2017-12-29 2018-04-10 Method for predicting processing rate of router based on minimum additive deconvolution Active CN108540323B (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201711472903X 2017-12-29
CN201711472903 2017-12-29

Publications (2)

Publication Number Publication Date
CN108540323A CN108540323A (en) 2018-09-14
CN108540323B true CN108540323B (en) 2020-02-21

Family

ID=63479696

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201810315283.7A Active CN108540323B (en) 2017-12-29 2018-04-10 Method for predicting processing rate of router based on minimum additive deconvolution

Country Status (1)

Country Link
CN (1) CN108540323B (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110362772B (en) * 2019-06-11 2022-04-01 北京邮电大学 Real-time webpage quality evaluation method and system based on deep neural network

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7680922B2 (en) * 2003-10-30 2010-03-16 Alcatel Lucent Network service level agreement arrival-curve-based conformance checking
CN101478456B (en) * 2009-01-16 2011-02-09 华中科技大学 Fast forwarding service end-to-end time delay prediction method
CN102780581B (en) * 2012-07-20 2014-10-22 北京航空航天大学 AFDX (Avionics Full Duplex Switched Ethernet) end-to-end delay bound claculation method based on random network calculus
CN105162612B (en) * 2015-06-17 2018-01-26 中国航空无线电电子研究所 A kind of overstocked method of data in calculating AFDX networks
CN105516206A (en) * 2016-01-28 2016-04-20 西南大学 Network intrusion detection method and system based on partial least squares

Also Published As

Publication number Publication date
CN108540323A (en) 2018-09-14

Similar Documents

Publication Publication Date Title
Liu et al. A new queueing model for QoS analysis of IEEE 802.11 DCF with finite buffer and load
CN101958833B (en) RED-based network congestion control algorithm
CN108880946B (en) Estimation method for data communication time delay between main station of wide area monitoring system and PMU (power management unit)
CN102780581B (en) AFDX (Avionics Full Duplex Switched Ethernet) end-to-end delay bound claculation method based on random network calculus
CN112733303B (en) Multi-strategy industrial TSN shaper modeling method based on deterministic network algorithm
CN107948103A (en) A kind of interchanger PFC control methods and control system based on prediction
Xu et al. On peak age of information in data preprocessing enabled IoT networks
Bakhouya et al. Analytical modeling and evaluation of on-chip interconnects using network calculus
CN108540323B (en) Method for predicting processing rate of router based on minimum additive deconvolution
Poojary et al. An asymptotic approximation for TCP CUBIC
CN101997776A (en) Router queue control system based on congestion identification and control method thereof
CN102970156A (en) Optimal modeling method for end-to-end delay in avionics full duplex switched Ethernet (AFDX) data networks
Zaidi et al. Voice packet performance estimation through step network using OPNET
Mahani et al. Heavy-tail and voice over internet protocol traffic: queueing analysis for performance evaluation
CN104092566B (en) Queue management method for service differentiation
Tao et al. Performance analysis on M2M communication network based on stochastic network calculus
Biddut et al. On the analysis of RED algorithm in ZigBee network for queue management
Cheng et al. Internet traffic characterization using packet-pair probing
Sharma et al. Performance analysis of TCP connections with RED control and exogenous traffic
Sunny et al. Delay modelling for a single-hop wireless mesh network under light aggregate traffic
Xu et al. A framework of efficient hybrid model and optimal control for multihop wireless networks
Jiang et al. Modelling delay propagation within a train communication network
Tolstikov et al. Data loss regulation to ensure information quality in sensor networks
Geyer et al. Towards stochastic flow-level network modeling: Performance evaluation of short TCP flows
Biddut et al. Queue management of RED enabled ZigBee network based on packet size variations and distribution techniques

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant