CN105162612A - Method for calculating data backlog in AFDX (Avionic Full-Duplex Switched Ethernet) network - Google Patents

Method for calculating data backlog in AFDX (Avionic Full-Duplex Switched Ethernet) network Download PDF

Info

Publication number
CN105162612A
CN105162612A CN201510336533.1A CN201510336533A CN105162612A CN 105162612 A CN105162612 A CN 105162612A CN 201510336533 A CN201510336533 A CN 201510336533A CN 105162612 A CN105162612 A CN 105162612A
Authority
CN
China
Prior art keywords
network
data
afdx
delay
overstock
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
Application number
CN201510336533.1A
Other languages
Chinese (zh)
Other versions
CN105162612B (en
Inventor
盛春玲
罗海
付佩
缪万胜
谷青范
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.)
China Aeronautical Radio Electronics Research Institute
Original Assignee
China Aeronautical Radio Electronics Research Institute
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 China Aeronautical Radio Electronics Research Institute filed Critical China Aeronautical Radio Electronics Research Institute
Priority to CN201510336533.1A priority Critical patent/CN105162612B/en
Publication of CN105162612A publication Critical patent/CN105162612A/en
Application granted granted Critical
Publication of CN105162612B publication Critical patent/CN105162612B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/14Network analysis or design
    • H04L41/142Network analysis or design using statistical or mathematical methods

Landscapes

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

Abstract

The invention discloses a method for calculating data backlog in an AFDX (Avionic Full-Duplex Switched Ethernet) network, which comprises the steps of S1, dividing the AFDX network into different subnets according to the priority, carrying out independent network analysis in allusion to each subset, and calculating an arrival curve and a service curve of each virtual link; S2, calculating data delay Di caused by queuing inside each switch; and S3, calculating data backlog of each node according to the data delay Di. Through the method disclosed by the invention, a lower data backlog upper bound in the AFDX network is acquired, thereby providing a more reliable theoretical basis and a larger configuration data optimization space for network resource distribution, and more effectively ensuring the certainty of the AFDX network.

Description

A kind of method calculating data in AFDX network and overstock
Technical field
The present invention relates to AFDX (AvionicFull-DuplexSwitchedEthernet) switch network analytical technology, particularly a kind of method calculating data in AFDX network and overstock in AFDX network deterministic parsing process.
Background technology
AFDX is the certainty airplane data networking bus system being applicable to avionics system information transmission.AFDX on Basic Ethernet, add certainty timing mechanism and reliable information transmission mechanisms is applied to adapt to avionics, certainty timing mechanism is realized by the virtual link technology with finite bandwidth and frame transmission interval, reliable information transmission is then realized by frame administrative mechanism, comprises redundant transmission and the integrity checking of information.The state of avionics system qualitative representation system is really foreseeable, namely knows that the current state of system and transition criterion just can the behaviors of Estimation System.Certainty is always correlated with safety analysis.In order to the safety of aircraft, deterministic requirement is applicable to each system level of aircraft.For data communication system, data communication must require to ensure certainty by network analysis.
Network calculus is a network certainty queuing theory, is the theoretical foundation of AFDX network deterministic parsing.The deterministic parsing of AFDX network needs Calculus Theory Network Based calculate delay, the shake of information transmission in worst case lower network and overstock.Network calculus is proposed in 1991 by L.Cruz, through the development of twenties years, especially the people such as Agrawal, LeBoudec is to the proposition of service curves, Min-plus algebra and class convolution concept, make network calculus become basis complete, be widely used, theory that system is huge.
The people such as C.Fraboul and J.Grieu are applied to AFDX network the basic theories of network calculus the earliest, propose than more complete model and algorithm.Wherein most important two parameters are exactly that data delay and data overstock.In engineering reality, the maximum data that network allows overstocks and is configured in units of frame number, therefore needs to calculate the upper bound of the frame number of the inner each priority backlog of switch in network transmission process in network analysis.Existing Mathematical Modeling and algorithm thereof are when calculated data overstocks the frame number upper bound, in units of byte, and all frames that tentation data buffering area is overstock are the maximum of the maximum frame length (MaxFrameSize) of virtual links (VL, VirtualLink) all in this priority.This computational methods are too coarse, and comparison of computational results is pessimistic, often overstocks the upper bound far beyond real data.
Summary of the invention
In order to the frame number computational methods that the data realizing more excellent AFDX network analysis modeling overstock, the invention provides a kind of method calculating data in AFDX network and overstock, obtain data lower in AFDX network and overstock the upper bound, for Resource Allocation in Networks provides more reliable theoretical foundation and larger configuration data to optimize space, more effectively ensure the certainty of AFDX network.
Goal of the invention of the present invention is achieved through the following technical solutions:
Calculate the method that in AFDX network, data overstock, comprise following steps:
Step one, AFDX network is divided into different subnets according to priority, independently carries out network analysis for each subnet, calculate the arrival curve of each virtual link in each switch internal damping exit position and service curves:
A) arrival curve of network in each subnet is calculated:
α(t)=σ+ρt,
Wherein, be σ burstiness, ρ is transmission rate;
B) service curves of network in each subnet is calculated:
β R,T(t)=R[t-T] +,
Wherein R is service speed, and T is service delay, symbol [x] +represent: if x>=0, this value equals x, otherwise this value is 0.
Step 2, the data delay D caused due to queuing by arrival curve and service curve calculation i, computing formula is:
D i=sup t≥0{inf τ≥0{α(t)≤β(t+τ)}}
Carry out calculated data without maximum transversal distance in step 3, the present invention and overstock the upper bound, but come the overstocked frame number upper bound of calculated data with the upper bound of data delay.According to data delay D ithe data calculating each node overstock:
Wherein: T tachnical delayfor the tachnical delay upper bound of each node, general T tachnical delay=50 μ s, T shapingfor from the tachnical delay be shaped into from starting terminal to export.
Accompanying drawing explanation
Fig. 1 is embodiment neutron network topology structure;
Fig. 2 is arrival curve schematic diagram in embodiment;
Fig. 3 is service curves schematic diagram in embodiment;
Fig. 4 is data delay schematic diagram under worst case in embodiment;
Fig. 5 is schematic flow sheet of the present invention.
Embodiment
In method in the past, it is after obtaining arrival curve and service curves that data delay and data overstock, get the horizontal computation of maximal distance data delay upper bound, get longitudinal computation of maximal distance data and overstock the upper bound, it is separately parallel computation that data delay and data overstock.Carry out calculated data without maximum transversal distance in the present invention and overstock the upper bound, but come the overstocked frame number upper bound of calculated data with the upper bound of data delay.
Due in AFDX network data send periodicity, within cycle regular hour, the frame number sent out from transmitting terminal is bounded above, the present invention according to every bar VL data in a network residence time computing network or certain node data overstock.
As shown in Figure 5, implementing procedure of the present invention is as follows:
Arrival curve and service curves is asked after step one, subregion.
Network partition:
First need network to be divided into different subnets according to priority, independently carry out network analysis calculating for each subnet.Namely for multipriority AFDX network, AFDX network need be divided into multiple subnet according to priority, each subnet only comprises the VL of same priority.Fig. 1 is the network topology structure figure of certain priority subnet.Wherein i represents network node (each outlet of switch has a buffering area, and each buffering area is a network node).
Such as, there is the VL of height (H), low (L) two priority in the AFDX network of A, B net redundancy, then can be divided into four subnets:
N (H, A): comprise all high priority VL netted by A;
N (H, B): comprise all high priority VL netted by B;
N (L, A): comprise all low priority VL netted by A;
N (L, B): comprise all low priority VL netted by A;
Calculate before data overstock in subnet, need first to calculate the arrival curve of network in this subnet, service curves and data delay.
Arrival curve:
VL is as the basic data transmission channel in AFDX network, and terminal is by controlling the maximum frame length S of Frame in every bar VL maxand send the transmission that interval BAG carrys out control information, according to the leakage prediction of shaping, the arrival curve of VL can be obtained:
α(t)=σ+ρt
Be wherein σ burstiness, ρ is transmission rate.Arrival curve schematic diagram as shown in Figure 2.
Service curves:
A basic service curves model is called rate delay service curves, as follows:
β R,T(t)=R[t-T] +
Wherein R is service speed, and T is service delay, symbol [x] +represent: if x>=0, this value equals x, otherwise this value is 0.
Service curves schematic diagram as shown in Figure 3.
Step 2, checking computations theory calculate Network Based go out the data delay of each virtual link.
By the network analysis of Calculus Theory Network Based, can calculate each VL inner due to the data delay caused of queuing up at each switch, we are designated as D this parameter i.
Assuming that data flow to into certain buffer joint, the arrival curve of this data flow is α (t), and service curves is β (t).The then maximum data delay D of this buffer joint i(unit: ms) is the maximum transversal distance between α (t) and β (t), as shown in Figure 4.Then there is following formula:
D i=sup t≥0{inf τ≥0{α(t)≤β(t+τ)}}。
Step 3, on the basis of data delay calculated data overstock.
1, every bar VL follows its transmission cycle, therefore timing from this point from transmitting terminal reshaper, assuming that from be shaped into from the tachnical delay starting terminal to export be T shaping, then every bar VL kdelay Bound before sending from transmitting terminal reshaper to certain node m for:
Illustrate:
A () set A comprises from transmitting terminal, VL kalong its sub-path arrive node m the set (comprising node m) that forms of all nodes of process.
B () transmitting time exceedes the frame of this time not in the buffering area of node m.
(c) T tachnical delay=50 μ s are the tachnical delay upper bound of each node.
2, in AFDX network, Frame periodically sends according to minimum bandwidth interval (BAG), then every bar VL k? the frame number upper bound sent in time is:
also the data being node m are overstocked, belong to VL kthe frame number upper bound, this result can be optimized further.
Represent from transmitting terminal by set B, VL karrive along its sub-path the previous node of node m the set (not comprising node m) that forms of all nodes of process.
Then there is following fundamental formular:
A the data of () m point overstock=and the data of set A overstock-and the data of set B overstock;
The data that b data that the data of () m point overstock the upper bound=set A overstock the upper bound-set B overstock next
And the Frame of not all overstocks at node i place, at VL kall nodes on path all can overstock, and the Nodes calculated below is before this minimum overstocked.
Assuming that entering node N mnode before is not queued up, and only there is tachnical delay.Due to
Total tachnical delay within this time is less than VL kone send the cycle, the minimum frame number that can overstock is 0.This illustrates in the time interval, on the path before node m, the lower bound that set B overstocks is 0.
Therefore, be just node N mthat overstocks belongs to VL kthe frame number upper bound of frame.
3, above calculating is all for a VL k, if having multiple through the VL of node m, so node m upper bound of all frame numbers of overstocking for:
4, the comprehensive formula of above 1,2,3, can obtain calculating the overstocked computational methods of AFDX network exchange machine data:
Be understandable that, for those of ordinary skills, can be equal to according to technical scheme of the present invention and inventive concept thereof and replace or change, and all these change or replace the protection range that all should belong to the claim appended by the present invention.

Claims (1)

1. calculate the method that in AFDX network, data overstock, comprise following steps:
Step one, AFDX network is divided into different subnets according to priority, independently carries out network analysis for each subnet, for each subnet, calculate arrival curve and the service curves of each virtual link according to the method for network checking computations;
Step 2 and then to calculate each virtual link inner due to the data delay D caused that queues up at each switch i;
Step 3, according to data delay D ithe data calculating each node overstock:
CN201510336533.1A 2015-06-17 2015-06-17 A kind of overstocked method of data in calculating AFDX networks Active CN105162612B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201510336533.1A CN105162612B (en) 2015-06-17 2015-06-17 A kind of overstocked method of data in calculating AFDX networks

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201510336533.1A CN105162612B (en) 2015-06-17 2015-06-17 A kind of overstocked method of data in calculating AFDX networks

Publications (2)

Publication Number Publication Date
CN105162612A true CN105162612A (en) 2015-12-16
CN105162612B CN105162612B (en) 2018-01-26

Family

ID=54803376

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201510336533.1A Active CN105162612B (en) 2015-06-17 2015-06-17 A kind of overstocked method of data in calculating AFDX networks

Country Status (1)

Country Link
CN (1) CN105162612B (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108540323A (en) * 2017-12-29 2018-09-14 西安电子科技大学 The method for predicting router processing speed based on minimum plus deconvolution
CN112733303A (en) * 2019-10-11 2021-04-30 中国科学院沈阳自动化研究所 Multi-strategy industrial TSN shaper modeling method based on deterministic network calculation
CN112821937A (en) * 2020-12-31 2021-05-18 广州大学 Data transmission method through satellite network, device and medium

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101894192A (en) * 2010-07-19 2010-11-24 北京航空航天大学 Simulation and demonstration system for design and validation of AFDX (Avionics Full Duplex Switched Ethernet) network and simulation and demonstration method thereof
CN102780581A (en) * 2012-07-20 2012-11-14 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
CN102970156A (en) * 2012-10-11 2013-03-13 上海交通大学 Optimal modeling method for end-to-end delay in avionics full duplex switched Ethernet (AFDX) data networks
US8958297B1 (en) * 2012-06-26 2015-02-17 Rockwell Collins, Inc. Data network with “per flow” flow monitoring

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101894192A (en) * 2010-07-19 2010-11-24 北京航空航天大学 Simulation and demonstration system for design and validation of AFDX (Avionics Full Duplex Switched Ethernet) network and simulation and demonstration method thereof
US8958297B1 (en) * 2012-06-26 2015-02-17 Rockwell Collins, Inc. Data network with “per flow” flow monitoring
CN102780581A (en) * 2012-07-20 2012-11-14 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
CN102970156A (en) * 2012-10-11 2013-03-13 上海交通大学 Optimal modeling method for end-to-end delay in avionics full duplex switched Ethernet (AFDX) data networks

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
HENRI BAUER: "《Worst-case backlog evaluation of Avionics switched Ethernet networks with the Trajectory approach》", 《IEEE 2012 24TH EUROMICRO CONFERENCE ON REAL-TIME SYSTEMS》 *
邱征: "《基于AFDX网络演算的性能分析技术研究》", 《电子世界》 *

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108540323A (en) * 2017-12-29 2018-09-14 西安电子科技大学 The method for predicting router processing speed based on minimum plus deconvolution
CN112733303A (en) * 2019-10-11 2021-04-30 中国科学院沈阳自动化研究所 Multi-strategy industrial TSN shaper modeling method based on deterministic network calculation
CN112733303B (en) * 2019-10-11 2024-03-12 中国科学院沈阳自动化研究所 Multi-strategy industrial TSN shaper modeling method based on deterministic network algorithm
CN112821937A (en) * 2020-12-31 2021-05-18 广州大学 Data transmission method through satellite network, device and medium

Also Published As

Publication number Publication date
CN105162612B (en) 2018-01-26

Similar Documents

Publication Publication Date Title
Scharbarg et al. A probabilistic analysis of end-to-end delays on an AFDX avionic network
US9363184B2 (en) Token bucket-based traffic limiting method and apparatus
CN102780581B (en) AFDX (Avionics Full Duplex Switched Ethernet) end-to-end delay bound claculation method based on random network calculus
Mifdaoui et al. Performance analysis of a master/slave switched ethernet for military embedded applications
Hua et al. Scheduling design and analysis for end-to-end heterogeneous flows in an avionics network
CN114422448A (en) Time-sensitive network traffic shaping method
CN105162612A (en) Method for calculating data backlog in AFDX (Avionic Full-Duplex Switched Ethernet) network
CN107070808A (en) A kind of the worst queueing delay processing method of the pre- constant current of AVB network multiclass based on credit shaping
CN112733303A (en) Multi-strategy industrial TSN shaper modeling method based on deterministic network calculation
Soni et al. Optimizing network calculus for switched ethernet network with deficit round robin
CN107682235A (en) A kind of propagation delay time prediction data dispatching method based on Kalman filtering
Mifdaoui et al. Full duplex switched ethernet for next generation" 1553B"-based applications
CN102970156A (en) Optimal modeling method for end-to-end delay in avionics full duplex switched Ethernet (AFDX) data networks
CN111079948B (en) SDN-based distributed machine learning training acceleration method
CN105119843A (en) A method for setting a message buffer queue length
Baga et al. Worst frame backlog estimation in an avionics full-duplex switched ethernet end-system
Jiang et al. Modelling delay propagation within a train communication network
Zimmermann et al. Towards reliability evaluation of AFDX avionic communication systems with rare-event simulation
EP2426874A1 (en) Bandwidth control method and bandwidth control device
Ge et al. Window adaptive backoff algorithm basing on statistical priority-based multiple access protocol
CN104486144B (en) A kind of AFDX network digitalizations end system deterministic simulation method
Stoltidis et al. Active Queue Management in Disaggregated 5G and Beyond Cellular Networks Using Machine Learning
Mifdaoui et al. Real-time characteristics of switched Ethernet for" 1553b"-embedded applications: Simulation and analysis
Ma et al. Mitigation of Sequence Inversion in AFDX based on Time-triggered Scheduling
CN110098967A (en) A kind of communication means, device, equipment and medium

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant