CN103986633B - Sub-network division method based on 1394b plurality of subnets transmission structures - Google Patents

Sub-network division method based on 1394b plurality of subnets transmission structures Download PDF

Info

Publication number
CN103986633B
CN103986633B CN201410222001.0A CN201410222001A CN103986633B CN 103986633 B CN103986633 B CN 103986633B CN 201410222001 A CN201410222001 A CN 201410222001A CN 103986633 B CN103986633 B CN 103986633B
Authority
CN
China
Prior art keywords
subnet
network
matrix
node
transmission
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.)
Expired - Fee Related
Application number
CN201410222001.0A
Other languages
Chinese (zh)
Other versions
CN103986633A (en
Inventor
伊小素
张量
王明
张春熹
宋凝芳
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.)
Beihang University
Original Assignee
Beihang 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 Beihang University filed Critical Beihang University
Priority to CN201410222001.0A priority Critical patent/CN103986633B/en
Publication of CN103986633A publication Critical patent/CN103986633A/en
Application granted granted Critical
Publication of CN103986633B publication Critical patent/CN103986633B/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Landscapes

  • Data Exchanges In Wide-Area Networks (AREA)

Abstract

The invention provides a kind of sub-network division method based on 1394b plurality of subnets transmission structures, belongs to network information communication technical field.This method step is:Obtain the matrix A of transmission bandwidth between network node;If network topology matrix be X, X's is often classified as a subnet, when r-th node is in certain subnet, to should subnet vector r-th row element value be 1;Enumerate all-network topological matrix;The delay matrix T in each subnet and between subnet under each network topology is calculated, the average delay of whole network is obtained;According to the corresponding network topology matrix of whole network average delay minima, sub-network division is carried out.Present invention achieves 1394b multi-subnet architectures are divided, overall system bandwidth is improve, system average delay is reduced, matrix description system topology is used, it is simple and easy, enormously simplify the calculating time during sub-network division.

Description

Sub-network division method based on 1394b plurality of subnets transmission structures
Technical field
The invention belongs to network information communication technical field, and in particular to a kind of based on 1394b plurality of subnets transmission structures Sub-network division method.
Background technology
IEEE1394 buses were developed in 1986 by Apple companies and formed standard in 1987.Initially Apple companies will Which is named as Firewire (live wire).FireWire specifications are by nineteen ninety-five electrical equipment and Electronic Engineering Association (IEEE) IEEE1394-1995;The agreement is further improved in IEEE in 2000 and release IEEE1394a;IEEE is based within 2002 IEEE1394-1995 and IEEE1394a are proposed IEEE1394b, and the agreement supports that fiber-optic transfer, bandwidth support 3.2Gbps, pole Improve greatly transmission performance.As the real-time of IEEE1394b buses is good, transfer rate is high, it is topological flexibly, extension easily and Hot plug is supported, becomes one of candidate's bussing technique of Aero-Space of future generation, industrial control field.
But 1394b buses are a kind of shared host-host protocols of traditional bus, its data by sharing bus transfer, Multi-channel data transmission cannot be realized.Although the country possesses some special knowledge to 1394b buses, with regard to 1394b multiplex techniques Research does not almost have.With the multiplex demand more and more higher of Large Volume Data in military field application, to 1394b transmission it is The research of the multiplex techniques of system is imperative.
It is different from traditional switching network, in 1394b plurality of subnets switching fabrics, port of the all-ones subnet by switch Group interconnection is so as to constitute whole data transmission system, and the node in each subnet passes through annular, chain, star-like or tree topology Shared 1394b fiber buss interconnected, as shown in Figure 1.Each subnet is mutual using traditional 1394b fiber buss It is coupled structure.When subnet internal node mutual data transmission, all-ones subnet is separate to carry out data transmission;When different sub-network When having data transfer between node, data are by source node of the switch from a subnet to the target section in another subnet Point.Accordingly, the data transfer of plurality of subnets switching fabric is divided into into two classes:Transmit between transmission and subnet in subnet.Son Transmission in net be traditional 1394b fiber buss transmission, this transmission means have the real-time of shared bussing technique with it is true It is qualitative.Transmit between subnet, realized by switch, its transmission performance is determined by switch.
IEEE was proposed 1394.1 agreements (1394.1 in 2004TM IEEE Standard for High Performance Serial Bus Bridges,Microprocessor and Microcomputer Standards Committee), the concept of 1394 bridges is which defined, as shown in Fig. 2 a plurality of 1394 bus can be constituted one by 1394 bridges The network of intercommunication, multiple bus can independently carry out data transmission and not interact, and data also can be passed by 1394 bridges It is defeated in other buses, realize the transmission of 1394b network parallel interaction datas.
Above-mentioned 1394 bridge architecture existing defects, when packet is transferred to by bridge in other buses, which will reach institute Have a bus, take the bandwidth of all buses, including being not intended to obtain the bus of the packet, so as to cause whole network efficiency compared with It is low.The agreement is not used widely, and does not also support the correlational study and product of the agreement.
The content of the invention
The present invention is based on for the problem that bus bandwidth occupancy is high during 1394b multiplexing, network transmission efficiency is relatively low A kind of queueing theory, it is proposed that sub-network division method based on 1394b plurality of subnets transmission structures.
The sub-network division method based on 1394b plurality of subnets transmission structures of the present invention, comprises the steps:
Step 1:Transmission bandwidth between network node is obtained, matrix A is expressed as follows:
N represents node number, the i-th row jth column element a in matrix AijRepresent node i roomy to the band that node j sends information It is little;
Step 2:If i-th subnet PiRepresent, then network topology matrix X=(P1,…,Pi,…,Pn)1×n, PiFor a n × 1 column vector, when certain subnet vector is vectorial for 0, represents that the subnet is not present;When r-th node is in subnet PiWhen interior, Then column vector PiR-th row element value be 1, be otherwise 0;
Step 3:The affiliated subnet of each node is set, all-network topological matrix is enumerated;
Step 4:To each network topology matrix, the time delay in each subnet and between subnet under the network topology is calculated, and is obtained Take the average delay of whole network;
Wherein, the time delay matrix T in each subnet and between subnet is represented:
Element T in matrix TiiRepresent subnet PiInterior time delay,μ be bus average service rate, λiiFor subnet PiInterior, subnet PiThe transmission bandwidth total amount between other subnets,
Element T in matrix TijRepresent subnet PiWith subnet PjBetween time delay, Tij=Tii+Tjj, TjjRepresent subnet PjInside prolong When;
The average delay t of whole network is:
Wherein, bbandFor the total bandwidth of network;
Step 5:According to the corresponding network topology matrix of whole network average delay minima, sub-network division is carried out.
Advantages of the present invention with good effect is:
(1) present invention achieves 1394b light bus multiplex techniques, are conducive to the application of 1394b light buses;
(2) the inventive method is based on 1394b plurality of subnets transmission structures, realizes the division of 1394b multi-subnet architectures, improves Overall system bandwidth, reduces system average delay;
(3) the inventive method is based on queueing theory, using matrix description system topology, simple and easy, enormously simplify The calculating time during sub-network division.
Description of the drawings
Fig. 1 is 1394b plurality of subnets switching fabric schematic diagrams;
Fig. 2 is 1394 bridge architecture schematic diagrams;
Fig. 3 is the schematic flow sheet of the sub-network division method of the present invention.
Specific embodiment
Below in conjunction with drawings and Examples, the present invention is described in further detail.
The present invention is directed to the multiplex problems of 1394b, according to 1394b plurality of subnets switching fabrics, it is proposed that a kind of simple Easily sub-network division method, by the transmission bandwidth and subnet topology structure of matrix description 1394b multiloop loop system interior joints, builds Vertical list window services model, calculate inside each subnet of plurality of subnets system, the bandwidth between subnet and time delay, selects time delay minimum Topological structure, realizes the division of 1394b multi-subnet architectures.The present invention can not only obtain system minimum time delay topological structure, real The sub-network division of existing 1394b plurality of subnets transmission structures, and the development of 1394b multiplex techniques will be promoted.The present invention is compared Originally the shared bus type transmission of 1394b traditional serial, realizes the transmission of 1394b multidiameter delay formulas, and it is total to improve system Bandwidth, reduces communication delay.
The invention provides a kind of sub-network division method based on 1394b plurality of subnets transmission structures, illustrates with reference to Fig. 3 Implement step.
If there is n node in the 1394b multiplex systems applied by the sub-network division method of the present invention, and based on following Setting:
(1) known to the data exchange amount of each node;
Different from the relatively random service condition of each node in family network, in the actual application background of avionics system Under, the inter-node transmission bandwidth of 1394b multi-subnet architectures is relative to be determined, can be evaluated whether its size.
(2) in plurality of subnets system, data exchange meets M/D/1 window schemes;
Based on queueing theory (TQ, the Theory of Queues), it is assumed that data are reached obeys negative the interval time of system Exponential, and be to determine the service time of each packet, then can be by data in subnet in plurality of subnets system, between subnet The mathematical model of M/D/1 window schemes is set up in transmission.M/D/1 window schemes are a simple queuing models, wherein M show to Up to Markov process is obeyed, D shows to leave being to determine and occur with fixed speed, and 1 shows only one server.
The plurality of subnets division methods of the present invention, the step of divide to the 1394b plurality of subnets topological structure of n node such as Under.
Step 1:Obtain transmission bandwidth between network node.
For the network with n node, all inter-node transmission amount of bandwidth are represented by the matrix A of n × n:
Wherein, the i-th row jth column element a in matrix AijRepresent the amount of bandwidth that node i sends information to node j.In network There is no situation of the node to itself transmission data, therefore aii=0.
Step 2:If network topology is expressed as matrix X, in matrix, each column vector represents a subnet.
For the network with n node, the plurality of subnets topological matrix of network is represented by the matrix X of n × n:
X=(P1,…,Pi,…,Pn)1×n (2)
Wherein, PiI-th subnet is represented, is the column vector of n × 1.When r-th node is in subnet PiWhen interior, then column vector PiR-th row element value be 1, be otherwise 0.When a certain subnet interior element is 0, expression does not have node in the subnet Interior, the subnet is not present.
Relatively, in network, each node, for subnet, can be represented with the row vector of a 1 × n, when r-th node exists Subnet PiWhen interior, the vector of corresponding node for (0 ... 0,1,0 ..., 0)1×n, wherein, r-th element value is 1, remaining element value For 0.Obtain for the available knot vector for belonging to the subnet of each subnet vector is added again transposition.If subnet PiSave comprising k Point, vectorial ε thereini,kRepresent subnet PiIn k-th node, vectorial Pi=(εi,1i,2+…εi,k)T
Step 3:The affiliated subnet of each node is set, all-network topological matrix is enumerated.
By arranging the affiliated subnet of each node, different network topologies are obtained.For the network of n node, Ke Nengcun The total n of network topology matrixnIt is individual, it is contemplated that network symmetry, half quantity (n can be calculated belown/ network topology square 2) Battle array.
Step 4:Determine the transmission bandwidth in each subnet and between each subnet under current network topology, and obtain whole network Average delay.
In network, between arbitrary node, amount of bandwidth is represented by:
εi、εjNode i, the vector of node j are represented respectively.
For the plurality of subnets system with n node, in each subnet, matrix of the transmission bandwidth size by n × n between subnet B is represented:
Wherein, biiEach node-node transmission total amount of bandwidth inside subnet i is represented, if there was only 1 node inside subnet or without section Point, biiFor 0;bijRepresent bandwidth sum of each node to each node-node transmission in subnet j in subnet i.
Through checking, inter-node transmission amount of bandwidth matrix A and its topological structure matrix X, transmit between each subnet, subnet The relation of bandwidth matrices B meets following formula:
B=XTAX (5)
Proof procedure:
If subnet PiIn have k node:i1,i2,…,ik, then subnet PiThe a width of any two of which of interior output transmission Internodal data transmission bandwidth sum:
Bring (3) formula into and can obtain subnet PiInterior data transfer bandwidth biiFor:
In the same manner, the data transfer bandwidth between subnet can be obtained.If subnet PiIn have m node:i1,i2,…,im, subnet Pj In have h node:j1,j2,…,jh, subnet PiWith subnet PjBetween data transfer bandwidth biiFor:
(7) and (8) two formulas are brought in formula (4),
That is B=XTAX。
Therefore, by the topological matrix X of the matrix A and plurality of subnets system of inter-node transmission amount of bandwidth, can obtain Transmission bandwidth matrix B in the subnet of 1394b plurality of subnets systems, between subnet.
In each subnet for n × n, between subnet in transmission bandwidth size matrix B, corresponding subnet, transmit between subnet and prolong When represented by the matrix T of n × n:
Wherein, TiiRepresent subnet PiInterior time delay.In subnet during data transfer, node calculates flat in a 1394b bus The waiting time needs the node (including that subnet is interior and between subnet) that data are sent in view of all and bus, according to M/D/1 Window scheme calculates average queuing latency formulaWherein λ is data packet number, and μ is bus average service Rate.For subnet Pi, λ is subnet PiInterior, subnet PiTransmission bandwidth total amount λ between other subnetsii
Time delayCan obtain.
In the same manner, TijFor subnet PiWith subnet PjBetween time delay.Data are by subnet PiInterior joint is transmitted to switch and is transferred to again Subnet PjInterior joint, as the time delay of switch was had been contemplated that in subnet internal delay time is calculated, so Tij=Tii+Tjj
Finally, the average delay t of acquisition whole network is
Wherein, bbandFor the total bandwidth of network.Formula (12) represents that the bandwidth ratio for taking each subnet is corresponding Delay-product be added, calculate the average delay of plurality of subnets system.
Step 5:According to the corresponding network topology matrix of whole network average delay minima, sub-network division is carried out.
The minimum topological matrix of whole network average delay is tminWhen corresponding XminMatrix, i.e.,

Claims (1)

1. a kind of sub-network division method based on 1394b plurality of subnets transmission structures, it is characterised in that comprise the steps:
Step 1:Transmission bandwidth between network node is obtained, matrix A is expressed as follows:
N represents node number, the i-th row jth column element a in matrix AijRepresent the amount of bandwidth that node i sends information to node j; There is no situation of the node to itself transmission data, therefore a in networkii=0;
Step 2:If i-th subnet PiRepresent, then network topology matrix X=(P1,…,Pi,…,Pn)1×n, PiFor n × 1 Column vector, when certain subnet vector is vectorial for 0, represents that the subnet is not present;When r-th node is in subnet PiWhen interior, then arrange Vectorial PiR-th row element value be 1, be otherwise 0;
Step 3:The affiliated subnet of each node is set, all-network topological matrix is enumerated;
Step 4:To each network topology matrix, the time delay in each subnet and between subnet under the network topology is calculated, and obtains whole The average delay of individual network;
Wherein, the time delay matrix T in each subnet and between subnet is represented:
Element T in matrix TiiRepresent subnet PiInterior time delay,μ be bus average service rate, λiiFor subnet PiInterior, Subnet PiThe transmission bandwidth total amount between other subnets,
Element T in matrix TijRepresent subnet PiWith subnet PjBetween time delay, Tij=Tii+Tjj, TjjRepresent subnet PjInterior time delay;
The average delay t of whole network is:
t = < X T A X b b a n d &CenterDot; T >
Wherein, bbandFor the total bandwidth of network;
Step 5:According to the corresponding network topology matrix of whole network average delay minima, sub-network division is carried out.
CN201410222001.0A 2014-05-23 2014-05-23 Sub-network division method based on 1394b plurality of subnets transmission structures Expired - Fee Related CN103986633B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201410222001.0A CN103986633B (en) 2014-05-23 2014-05-23 Sub-network division method based on 1394b plurality of subnets transmission structures

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201410222001.0A CN103986633B (en) 2014-05-23 2014-05-23 Sub-network division method based on 1394b plurality of subnets transmission structures

Publications (2)

Publication Number Publication Date
CN103986633A CN103986633A (en) 2014-08-13
CN103986633B true CN103986633B (en) 2017-03-29

Family

ID=51278460

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201410222001.0A Expired - Fee Related CN103986633B (en) 2014-05-23 2014-05-23 Sub-network division method based on 1394b plurality of subnets transmission structures

Country Status (1)

Country Link
CN (1) CN103986633B (en)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110838962B (en) * 2019-11-18 2021-11-12 上海船舶研究设计院(中国船舶工业集团公司第六0四研究院) Marine network system and marine system
CN112612224A (en) * 2020-12-04 2021-04-06 中国航空工业集团公司成都飞机设计研究所 Portable 1394 bus topology customization and expansion device

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101795262A (en) * 2009-12-09 2010-08-04 北京航空航天大学 IEEE-1394b bus and CAN bus protocol converter based on microprocessor
CN102546336A (en) * 2011-12-28 2012-07-04 北京航空航天大学 IEEE (Institute of Electrical and Electronics Engineers)-1394b optical bus protocol converter based on Versa PHY (Physical Layer)

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101795262A (en) * 2009-12-09 2010-08-04 北京航空航天大学 IEEE-1394b bus and CAN bus protocol converter based on microprocessor
CN102546336A (en) * 2011-12-28 2012-07-04 北京航空航天大学 IEEE (Institute of Electrical and Electronics Engineers)-1394b optical bus protocol converter based on Versa PHY (Physical Layer)

Also Published As

Publication number Publication date
CN103986633A (en) 2014-08-13

Similar Documents

Publication Publication Date Title
US10454585B2 (en) Data center network system and signal transmission system
CN105162721B (en) Full light network data centre network system and data communications method based on software defined network
CN104429029B (en) Being controlled based on the service rate for postponing by central controller in network
Chen et al. WaveCube: A scalable, fault-tolerant, high-performance optical data center architecture
CN108566659B (en) 5G network slice online mapping method based on reliability
CN102882811B (en) Based on data center light interconnection network system and the communication means of array waveguide grating
CN106254254B (en) Mesh topology structure-based network-on-chip communication method
CN103346979B (en) Flow distribution method and equipment in SPBM network
CN103428805B (en) The virtual mapping method of a kind of wireless network based on link anti-interference
CN105022717A (en) Network on chip resource arbitration method and arbitration unit of additional request number priority
JP5521258B2 (en) Cluster router and cluster routing method
CN104049533A (en) Network control system controller design method based on hidden semi-markov model
CN111049667B (en) Time-triggered Ethernet communication service offline scheduling optimization method
CN105072040B (en) Efficient satellite SpaceWire plug and play network topological construction methods
CN103986633B (en) Sub-network division method based on 1394b plurality of subnets transmission structures
CN111193971A (en) Machine learning-oriented distributed computing interconnection network system and communication method
CN103200468B (en) The route Wavelength allocation method of power optical fiber communication network and device
CN104081693B (en) Optics connects reconfiguring for infrastructure
CN107070681A (en) Network topology acquisition methods and device based on software defined network SDN
CN106789750B (en) A kind of high-performance calculation interconnected network system and communication means
Zhao et al. Unified control system for heterogeneous networks with Software Defined Networking (SDN)
CN103546949B (en) A kind of Access Network bandwidth scheduling method based on dormancy power-saving mechanism
CN103001869B (en) Method and system of cloud route data processing based on port deflection
CN105337872A (en) Control plane network partitioning method based on energy efficiency precedence
CN110445574A (en) A kind of optical network transmission method and system based on hypergraph structure

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
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20170329

Termination date: 20200523

CF01 Termination of patent right due to non-payment of annual fee