CN106302221A - Traffic scheduling method based on end office's cloud and system - Google Patents
Traffic scheduling method based on end office's cloud and system Download PDFInfo
- Publication number
- CN106302221A CN106302221A CN201610819204.7A CN201610819204A CN106302221A CN 106302221 A CN106302221 A CN 106302221A CN 201610819204 A CN201610819204 A CN 201610819204A CN 106302221 A CN106302221 A CN 106302221A
- Authority
- CN
- China
- Prior art keywords
- port
- flow
- vcpe
- network
- load data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- 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
- H04L47/12—Avoiding congestion; Recovering from congestion
- H04L47/125—Avoiding congestion; Recovering from congestion by balancing the load, e.g. traffic engineering
-
- 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
- H04L47/24—Traffic characterised by specific attributes, e.g. priority or QoS
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L47/00—Traffic control in data switching networks
- H04L47/50—Queue scheduling
- H04L47/52—Queue scheduling by attributing bandwidth to queues
- H04L47/527—Quantum based scheduling, e.g. credit or deficit based scheduling or token bank
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L47/00—Traffic control in data switching networks
- H04L47/50—Queue scheduling
- H04L47/62—Queue scheduling characterised by scheduling criteria
- H04L47/625—Queue scheduling characterised by scheduling criteria for service slots or service orders
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The present invention provides a kind of traffic scheduling method based on end office's cloud and system, and wherein, the method includes: vCPE obtains the flow load data of each port;The flow load data of each port that vCPE will get, store in the first storage matrix;VCPE is optimized lexical analysis to the flow load data of each port in the first storage matrix, to determine the Optimized Operation strategy of each port;The flow of each port, according to the Optimized Operation strategy of each port, is scheduling by vCPE.New adjustment decision-making is obtained according to real-time each port flow state, determine the port of imbalance, the port of imbalance is adjusted, the each port i.e. network traffics of node are made to carry out overall situationization Optimized Operation, achieve the floating resources of each port can reach optimization, the response service time can the least, network traffics throughput can be the biggest, and then reduce network congestion degree, improve network utilization, reduce consumption.
Description
Technical field
The present invention relates to communication technical field, particularly relate to a kind of traffic scheduling method based on end office's cloud and system.
Background technology
End office's cloudization has become the important trend of global telecommunications industry development, and end office's cloud technology is existing in order to solve
The deficiency of specialized communication equipment and produce therewith.Its function can be aggregated or decompose, and instantiation in infrastructure
The function assembled or decompose.Currently, end office's cloud has become as the development priority of global telecommunications industrial circle.Existing end office (EO) cloud
Application system is along with the quick growth of telecommunications applied business, and the flow in network increases rapidly, need flow is scheduling and
Control.The problems such as the network congestion degree produced of companion is high, utilization rate is low, consumption is high become increasingly conspicuous.
In prior art, end office's cloud application system mainly uses static set Chinese style flow scheduling mode, goes to carry out flow
Scheduling, static set Chinese style flow scheduling mode has connection-oriented characteristic, by QoS negotiation, call access control, reservation bandwidth etc.
Mechanism realizes scheduling and the control of flow.
But in prior art, the QoS negotiation of employing, call access control, leave strip in static set Chinese style flow scheduling mode
The mechanism such as width, still cannot well solve the problem that network congestion degree is high, and network utilization is low, consume height.
Summary of the invention
The present invention provides a kind of traffic scheduling method based on end office's cloud and system, in order to solve cannot in prior art
Well solve the problem that network congestion degree is high, and network utilization is low, consume high problem.
It is an aspect of the present invention to provide a kind of traffic scheduling method based on end office's cloud, including:
VCPE obtains the flow load data of each port;
The flow load data of each port that described vCPE will get, store in the first storage matrix;
Described vCPE is optimized lexical analysis to the flow load data of each port in the first storage matrix, to determine
The Optimized Operation strategy of each port;
The flow of each port, according to the Optimized Operation strategy of each port, is scheduling by described vCPE.
Another aspect of the present invention is to provide a kind of flow scheduling system based on end office's cloud, including:
VCPE and each port;
Described vCPE, for obtaining the flow load data of each port;The flow load data of each port that will get,
Store in the first storage matrix;The flow load data of each port in the first storage matrix are optimized lexical analysis,
To determine the Optimized Operation strategy of each port;Optimized Operation strategy according to each port, is scheduling the flow of each port.
The solution have the advantages that: obtain, by OS Server, the port flow load data that each port obtains, will obtain
The port flow load data of each port taken, stores in the first storage matrix;According to user gradation optimization priority scheduling
Evaluation function, determines flow, time delay and the port of packet loss imbalance;The port of imbalance is touched by OS Server submodule to be used
Family level priority Optimization scheduling algorithm, dynamically adjusts the port of imbalance, enters the data mapping matrix of the port of imbalance
Row sum-equal matrix.It is thus possible to obtain new adjustment decision-making according to real-time each port flow state, determine the port of imbalance, to mistake
The port adjusted is adjusted so that each port i.e. network traffics of node carry out overall situationization Optimized Operation, it is achieved that Mei Geduan
Mouthful floating resources can reach optimization, the response service time can the least, network traffics throughput can to the greatest extent may be used
Can big, and then reduce network congestion degree, improve network utilization, reduce consumption.
Accompanying drawing explanation
The flow chart of the traffic scheduling method based on end office's cloud that Fig. 1 provides for the embodiment of the present invention one;
The flow chart of the traffic scheduling method based on end office's cloud that Fig. 2 provides for the embodiment of the present invention one;
The structural representation of the flow scheduling system based on end office's cloud that Fig. 3 provides for the embodiment of the present invention three;
The structural representation of the flow scheduling system based on end office's cloud that Fig. 4 provides for the embodiment of the present invention four.
Detailed description of the invention
For making the purpose of the embodiment of the present invention, technical scheme and advantage clearer, below in conjunction with the embodiment of the present invention
In accompanying drawing, the technical scheme in the embodiment of the present invention is clearly and completely described, it is clear that described embodiment is
The a part of embodiment of the present invention rather than whole embodiments.Based on the embodiment in the present invention, those of ordinary skill in the art
The every other embodiment obtained under not making creative work premise, broadly falls into the scope of protection of the invention.
The flow chart of the traffic scheduling method based on end office's cloud that Fig. 1 provides for the embodiment of the present invention one, such as Fig. 1 institute
Showing, the method for the present embodiment includes:
Step 101, vCPE obtain the flow load data of each port.
Wherein, the specific implementation of step 101 is:
Each port actively reports the flow load data of acquisition every preset time to vCPE;Or, vCPE periodically inquires
The flow load data that each port obtains.
In the present embodiment, concrete, VPN server is linked into vBRAS and is authenticated accessing internal enterprise resources,
It is single that the terminal of fixed network enterprises and institutions user accesses internal enterprise resources such as enterprises and institutions by the PG of addressable internal enterprise resources
The internal office work system of position.In this process, having OS Server submodule on vCPE, each vBRAS module has one
Individual OS Client submodule.Each OS Client submodule actively reports flow, OS Server to OS Server submodule
Module automated periodic uses flow scheduling strategy to carry out flow scheduling.
Each port actively reports the flow load data of acquisition, or OS every preset time to OS Server submodule
Server submodule periodically inquires the flow load data that each port obtains, and wherein, flow load data include port flow
Load, time delay and packet loss data.
The flow load data of each port that step 102, vCPE will get, store in the first storage matrix.
In the present embodiment, concrete, the port flow of each port obtained is loaded by the OS Server submodule of vCPE
Data, store in the first storage matrix.
Step 103, vCPE are optimized lexical analysis to the flow load data of each port in the first storage matrix, with
Determine the Optimized Operation strategy of each port.
In the present embodiment, concrete, the OS Server submodule of vCPE is for the stream of port each in the first storage matrix
Amount load data, is optimized lexical analysis, determines the Optimized Operation strategy of each port.
The flow of each port, according to the Optimized Operation strategy of each port, is scheduling by step 104, vCPE.
Wherein, the specific implementation of step 104 is: vCPE according to user gradation optimization priority scheduling evaluation function,
Determine flow, time delay and the port of packet loss imbalance;VCPE carries out flow optimization and adjustment to the port of imbalance.
VCPE, according to user gradation optimization priority scheduling evaluation function, determines flow, time delay and packet loss imbalance
Port, including:
Determine user gradation optimization priority scheduling evaluation function Wherein, xij kFor port whether load traffic definition value and π, ν ∈ (0,1), Qij kAnd Vij kIt is respectively port xij kCongestion Level SPCC based on its priority satisfied
Degree and flow availability, ε andIt is respectively Congestion Level SPCC satisfaction and the Dynamic gene of flow availability;K represents that kth time is repeatedly
Generation, k≤d and k=1,2 ..., d;PRIij k、delayij k、LRij kAnd FLOWij kIt is respectively port priority, port time delay, port
Packet loss and port flow, θ, σ, π and ν are Dynamic gene;
Determine that the first storage matrix after adjustment is Mij k+1=vij k+1*tu, wherein, vij k+1=τij k*vij k+ρ*(MPk-
Mij k)+ξ*(MGk-Mij k),τij kFor autonomic learning pheromone concentration weight, vij kFor kth time iteration
Middle Mij kDirection vector, tuFor unit time quantum, vij k+1For the direction vector of+1 iteration of kth, ρ and ξ be adjustment because of
Son, MPkAnd MGkIt is respectively in kth time iteration the global port of k iteration before current iteration local optimum port data vector sum
Data vector;Mij kIt is the first storage matrix, Mij k={ delayij k,LRij k,FLOWij k};
When determining that user gradation optimization priority scheduling evaluation function is not satisfied, or current iteration number of times is less than or equal to
During maximum iteration time, it is determined that the flow imbalance of present port, wherein, user gradation optimization priority scheduling evaluation function is not
It is satisfied the value for user gradation optimization priority scheduling evaluation function less than or equal to the user gradation of each port in iteration first
The meansigma methods of the value of optimization priority scheduling evaluation function.
In the present embodiment, concrete, the OS Server submodule of vCPE is according to the Optimized Operation strategy of each port, right
The flow of each port is scheduling.
Wherein, OS Server submodule, for the flow load data of port each in the first storage matrix, is optimized tune
Degree is analyzed, and determines the Optimized Operation strategy of each port;Then, OS Server submodule is according to the Optimized Operation plan of each port
Slightly, the flow of each port is scheduling, specifically includes: OS Server submodule is according to user gradation optimization priority scheduling
Evaluation function, determines flow, time delay and the port of packet loss imbalance;The port of imbalance is flowed by OS Server submodule
Amount optimizes and adjusts.OS Server submodule according to user gradation optimization priority scheduling evaluation function, determine flow, time
Prolong and the port of packet loss imbalance, include: arranging iteration maximum algebraically d is 200, and according to the first storage matrix, initialize
Port data mapping matrix;M in port data mapping matrixij kFor comprising port time delay, the packet loss of port and port flow
The port data vector of three aspects, is Mij k={ delayij k,LRij k,FLOWij k};Wherein, matrix Mij kAs follows:
The definition value of port whether load traffic isUser gradation optimization is preferentially adjusted
Degree evaluation function isε、Wherein, And θ, σ ∈ (0,1),π, ν ∈ (0,1), Qij kAnd Vij kIt is respectively port xij kCongestion Level SPCC based on its priority satisfied
Degree and flow availability, ε andBeing respectively Congestion Level SPCC satisfaction and the Dynamic gene of flow availability, k represents that kth time is repeatedly
In generation, wherein k must is fulfilled for k≤d condition, need to meet k=1, and 2 ..., the condition of d, PRIij k、delayij k、LRij kAnd FLOWij kPoint
Port priority, port time delay, the packet loss of port and port flow θ, σ, π and ν Wei not be Dynamic gene.After adjustment
One storage matrix is Mij k+1=vij k+1*tu, wherein, vij k+1=τij k*vij k+ρ*(MPk-Mij k)+ξ*(MGk-Mij k),τij kFor autonomic learning pheromone concentration weight, vij kFor M in kth time iterationij kDirection vector, tu
For unit time quantum, vij k+1For the direction vector of+1 iteration of kth, ρ and ξ is Dynamic gene, MPkAnd MGkIt is respectively kth
The global port data vector of k iteration (containing this) before current iteration local optimum port data vector sum in secondary iteration;When
Determine that user gradation optimization priority scheduling evaluation function is not satisfied, or current iteration number of times is less than or equal to greatest iteration time
Number time, it is determined that present port flow imbalance, wherein, user gradation optimization priority scheduling evaluation function be not satisfied into
The value of family grade optimization priority scheduling evaluation function is less than or equal to, and in iteration, the user gradation optimization of each port is preferential first
The meansigma methods of the value of scheduling evaluation function;The preferential Optimization scheduling algorithm of user gradation can be touched, the port of imbalance is moved
State adjusts, and is adjusted the data mapping matrix of the port of imbalance, carries out global optimization with the port optimized and revised by needs,
To realize the overall dynamics optimization of each port flow load;According to the data mapping matrix of the port after adjusting, adjust each port
Data, and will adjust after each port data be mapped as the second storage matrix.
The present embodiment obtains, by OS Server, the port flow load data that each port obtains, each port that will obtain
Port flow load data, store in the first storage matrix;According to user gradation optimization priority scheduling evaluation function, really
Make flow, time delay and the port of packet loss imbalance;It is the most excellent that the port of imbalance is touched user gradation by OS Server submodule
Change dispatching algorithm, the port of imbalance is dynamically adjusted, the data mapping matrix of the port of imbalance is adjusted.It is thus possible to
Enough obtain new adjustment decision-making according to real-time each port flow state, determine the port of imbalance, the port of imbalance is carried out
Adjust so that each port i.e. network traffics of node carry out overall situationization Optimized Operation, it is achieved that the floating resources of each port
Can reach optimization, the response service time can the least, network traffics throughput can be big as far as possible, and then fall
Low network congestion degree, improves network utilization, reduces consumption.
The flow chart of the traffic scheduling method based on end office's cloud that Fig. 2 provides for the embodiment of the present invention one, in embodiment
On the basis of one, as in figure 2 it is shown, the method for the present embodiment, before step 101, also include:
Step 201, the terminal of fixed network domestic consumer initiate DHCP request message;Fixing end by PG, Vxlan gateway and
VCPE sets up Vxlan tunnel, to complete the access of Vxlan double layer network;The DHCP request message of user USER_PC passes through
Vxlan double layer network is pass-through to vCPE;VCPE supports as DHCP Server to the terminal distribution IP address of lower extension, and supports
Process and send to vBRAS through fire wall etc. as IPoE/PPPoE Client and carry out dial-up access certification;NAT carries out private
Net and the conversion of public network address;NAT is by gateway accessing and accesses Internet;PG by the Vxlan gateway set up and
Vxlan tunnel between vCPE, is linked into addressable internal enterprise resources by Vxlan double layer network;Fixed network enterprises and institutions
The terminal of user accesses the internal office work system of internal enterprise resources such as enterprises and institutions by the PG of addressable internal enterprise resources
System;
Or, mobile terminal device initiates to access the message request of internal enterprise resources, and accesses telecommunications fortune by base station
The 3G/4G network of battalion business;Mobile terminal device passes through 3G/4G network insertion VPN server;Mobile terminal device passes through VPN
Server is linked into vBRAS and is authenticated accessing internal enterprise resources.
In the present embodiment, concrete, use loose coupling framework, its general frame is from being laterally divided into two territories: mobile
Terminal unit access domain and fixed network enterprises and institutions user's access domain.For fixed network enterprises and institutions user's access domain, the most solid
Net domestic consumer USER_PC initiates DHCP request message;Fixing end sets up Vxlan tunnel by PG, Vxlan gateway and vCPE,
And then realize the access of Vxlan double layer network;The request message of user USER_PC is pass-through to vCPE by Vxlan double layer network;
VCPE supports as DHCP Server to the terminal distribution IP address of lower extension, and supports to pass through as IPoE/PPPoE Client
Fire walls etc. process and send to vBRAS and carry out dial-up access certification;The conversion of private network and public network address is carried out by NAT;NAT leads to
Cross gateway accessing and access Internet;PG passes through the Vxlan tunnel between Vxlan gateway and the vCPE set up, and passes through
Vxlan double layer network is linked into addressable internal enterprise resources;The terminal of fixed network enterprises and institutions user is by addressable enterprise
The PG of internal resource accesses the internal office work system of internal enterprise resources such as enterprises and institutions.Mobile terminal device is accessed
Territory, mobile terminal device is initiated to access the message request of internal enterprise resources, and is accessed the 3G/4G of telecom operators by base station
Network;The mobile terminal device 3G/4G network insertion VPN server by telecom operators;Then mobile terminal device passes through
VPN server is linked into vBRAS and is authenticated accessing internal enterprise resources.
The present embodiment is linked into vBRAS by mobile terminal device by VPN server and is authenticated accessing in enterprise
Portion's resource, the terminal of fixed network enterprises and institutions user accesses internal enterprise resources such as enterprise by the PG of addressable internal enterprise resources
The internal office work system of public institution.In this process, OS Server obtains the port flow load number that each port obtains
According to, by the port flow load data of each port of acquisition, store in the first storage matrix;Excellent according to user gradation optimization
First dispatch evaluation function, determine flow, time delay and the port of packet loss imbalance;The OS Server submodule port to imbalance
Touch the preferential Optimization scheduling algorithm of user gradation, the port of imbalance is dynamically adjusted, the data of the port of imbalance are mapped
Matrix is adjusted.It is thus possible to obtain new adjustment decision-making according to real-time each port flow state, determine the end of imbalance
Mouthful, the port of imbalance is adjusted so that each port i.e. network traffics of node carry out overall situationization Optimized Operation, it is achieved that
The floating resources of each port can reach optimization, the response service time can the least, network traffics throughput energy
Enough big as far as possible, and then reduce network congestion degree, improve network utilization, reduce consumption.
The structural representation of the flow scheduling system based on end office's cloud that Fig. 3 provides for the embodiment of the present invention three, such as Fig. 3
Shown in, the system that the present embodiment provides, including:
VCPE31 and each port 32;
VCPE31, for obtaining the flow load data of each port 32;The flow load number of each port 32 that will get
According to, store in the first storage matrix;The flow load data of each port 32 in the first storage matrix are optimized scheduling
Analyze, to determine the Optimized Operation strategy of each port 32;According to the Optimized Operation strategy of each port 32, the flow to each port 32
It is scheduling.
Each port 32, specifically for: actively report the flow load data of acquisition to vCPE31 every preset time;Or
Person, vCPE31, specifically for: periodically inquire the flow load data that each port 32 obtains.
VCPE31, comprises determining that module 311, for according to user gradation optimization priority scheduling evaluation function, determines
Flow, time delay and the port of packet loss imbalance;Adjusting module 312, for carrying out flow optimization and adjustment to the port of imbalance.
Wherein, adjusting module 312, specifically for:
Determine user gradation optimization priority scheduling evaluation function
ε、Wherein, xij kFor port whether load traffic definition value and π, ν ∈ (0,1), Qij kAnd Vij kIt is respectively port xij kCongestion Level SPCC based on its priority satisfied
Degree and flow availability, ε andIt is respectively Congestion Level SPCC satisfaction and the Dynamic gene of flow availability;K represents that kth time is repeatedly
Generation, k≤d and k=1,2 ..., d;PRIij k、delayij k、LRij kAnd FLOWij kIt is respectively port priority, port time delay, port
Packet loss and port flow, θ, σ, π and ν are Dynamic gene;
Determine that the first storage matrix after adjustment is Mij k+1=vij k+1*tu, wherein, vij k+1=τij k*vij k+ρ*(MPk-
Mij k)+ξ*(MGk-Mij k),τij kFor autonomic learning pheromone concentration weight, vij kFor kth time iteration
Middle Mij kDirection vector, tuFor unit time quantum, vij k+1For the direction vector of+1 iteration of kth, ρ and ξ be adjustment because of
Son, MPkAnd MGkIt is respectively in kth time iteration the global port of k iteration before current iteration local optimum port data vector sum
Data vector;Mij kIt is the first storage matrix, Mij k={ delayij k,LRij k,FLOWij k};
When determining that user gradation optimization priority scheduling evaluation function is not satisfied, or current iteration number of times is less than or equal to
During maximum iteration time, it is determined that the flow imbalance of present port, wherein, user gradation optimization priority scheduling evaluation function is not
It is satisfied the value for user gradation optimization priority scheduling evaluation function less than or equal to the user gradation of each port in iteration first
The meansigma methods of the value of optimization priority scheduling evaluation function.
The traffic scheduling method based on end office's cloud of the present embodiment can perform the embodiment of the present invention one provide based on end
The traffic scheduling method of office's cloud, it is similar that it realizes principle, and here is omitted.
The present embodiment obtains, by OS Server, the port flow load data that each port obtains, each port that will obtain
Port flow load data, store in the first storage matrix;According to user gradation optimization priority scheduling evaluation function, really
Make flow, time delay and the port of packet loss imbalance;It is the most excellent that the port of imbalance is touched user gradation by OS Server submodule
Change dispatching algorithm, the port of imbalance is dynamically adjusted, the data mapping matrix of the port of imbalance is adjusted.It is thus possible to
Enough obtain new adjustment decision-making according to real-time each port flow state, determine the port of imbalance, the port of imbalance is carried out
Adjust so that each port i.e. network traffics of node carry out overall situationization Optimized Operation, it is achieved that the floating resources of each port
Can reach optimization, the response service time can the least, network traffics throughput can be big as far as possible, and then fall
Low network congestion degree, improves network utilization, reduces consumption.
The structural representation of the flow scheduling system based on end office's cloud that Fig. 4 provides for the embodiment of the present invention four, in reality
On the basis of executing example three, as shown in Figure 4, the system that the present embodiment provides, also include:
The terminal 41 of fixed network domestic consumer and mobile terminal device 42;
The terminal 41 of fixed network domestic consumer, before obtain the flow load data of each port 32 at vCPE31, initiates
DHCP request message, so that fixing end sets up Vxlan tunnel by PG, Vxlan gateway and vCPE31, to complete Vxlan bis-layers
The access of network;The DHCP request message of user USER_PC is pass-through to vCPE31 by Vxlan double layer network;VCPE31 supports
Give the terminal distribution IP address of lower extension as DHCP Server, and support as IPoE/PPPoE Client through fire wall etc.
Process and send to vBRAS and carry out dial-up access certification;NAT carries out the conversion of private network and public network address;NAT passes through gateway accessing
And access Internet;PG passes through the Vxlan tunnel between Vxlan gateway and the vCPE31 set up, by bis-layers of net of Vxlan
Network is linked into addressable internal enterprise resources;The terminal 41 of fixed network domestic consumer is accessed by the PG of addressable internal enterprise resources
The internal office work system of internal enterprise resources such as enterprises and institutions;
Mobile terminal device 42, before obtain the flow load data of each port 32 at vCPE31, initiates to access enterprise
The message request of industry internal resource, and the 3G/4G network of telecom operators is accessed by base station;Mobile terminal device 42 passes through
3G/4G network insertion VPN server;Mobile terminal device is linked into vBRAS by VPN server and is authenticated accessing enterprise
Industry internal resource.
The traffic scheduling method based on end office's cloud of the present embodiment can perform the embodiment of the present invention two provide based on end
The traffic scheduling method of office's cloud, it is similar that it realizes principle, and here is omitted.
The present embodiment obtains, by OS Server, the port flow load data that each port obtains, each port that will obtain
Port flow load data, store in the first storage matrix;According to user gradation optimization priority scheduling evaluation function, really
Make flow, time delay and the port of packet loss imbalance;It is the most excellent that the port of imbalance is touched user gradation by OS Server submodule
Change dispatching algorithm, the port of imbalance is dynamically adjusted, the data mapping matrix of the port of imbalance is adjusted.It is thus possible to
Enough obtain new adjustment decision-making according to real-time each port flow state, determine the port of imbalance, the port of imbalance is carried out
Adjust so that each port i.e. network traffics of node carry out overall situationization Optimized Operation, it is achieved that the floating resources of each port
Can reach optimization, the response service time can the least, network traffics throughput can be big as far as possible, and then fall
Low network congestion degree, improves network utilization, reduces consumption.
One of ordinary skill in the art will appreciate that: all or part of step realizing above-mentioned each method embodiment can be led to
The hardware crossing programmed instruction relevant completes.Aforesaid program can be stored in a computer read/write memory medium.This journey
Sequence upon execution, performs to include the step of above-mentioned each method embodiment;And aforesaid storage medium includes: ROM, RAM, magnetic disc or
The various media that can store program code such as person's CD.
Last it is noted that above example is only in order to illustrate technical scheme, it is not intended to limit;Although
With reference to previous embodiment, the present invention is described in detail, it will be understood by those within the art that: it still may be used
So that the technical scheme described in foregoing embodiments to be modified, or wherein portion of techniques feature is carried out equivalent;
And these amendment or replace, do not make appropriate technical solution essence depart from various embodiments of the present invention technical scheme spirit and
Scope.
Claims (10)
1. a traffic scheduling method based on end office's cloud, it is characterised in that including:
VCPE obtains the flow load data of each port;
The flow load data of each port that described vCPE will get, store in the first storage matrix;
Described vCPE is optimized lexical analysis to the flow load data of each port in the first storage matrix, to determine each end
The Optimized Operation strategy of mouth;
The flow of each port, according to the Optimized Operation strategy of each port, is scheduling by described vCPE.
Method the most according to claim 1, it is characterised in that described vCPE obtains the flow load data of each port, bag
Include:
Each port actively reports the flow load data of acquisition every preset time to described vCPE;
Or, described vCPE periodically inquires the flow load data that each port obtains.
Method the most according to claim 1, it is characterised in that described vCPE is according to the Optimized Operation strategy of each port, right
The flow of each port is scheduling, including:
Described vCPE, according to user gradation optimization priority scheduling evaluation function, determines flow, time delay and packet loss imbalance
Port;
Described vCPE carries out flow optimization and adjustment to the port of imbalance.
Method the most according to claim 3, it is characterised in that described vCPE comments according to user gradation optimization priority scheduling
Valency function, determines flow, time delay and the port of packet loss imbalance, including:
Determine user gradation optimization priority scheduling evaluation function Wherein, xij kFor port whether load traffic definition value and π, ν ∈ (0,1), Qij kAnd Vij kIt is respectively port xij kCongestion Level SPCC based on its priority satisfied
Degree and flow availability, ε andIt is respectively Congestion Level SPCC satisfaction and the Dynamic gene of flow availability;K represents that kth time is repeatedly
Generation, k≤d and k=1,2 ..., d;PRIij k、delayij k、LRij kAnd FLOWij kIt is respectively port priority, port time delay, port
Packet loss and port flow, θ, σ, π and ν are Dynamic gene;
Determine that the first storage matrix after adjustment is Mij k+1=vij k+1*tu, wherein, vij k+1=τij k*vij k+ρ*(MPk-Mij k)+ξ*
(MGk-Mij k),τij kFor autonomic learning pheromone concentration weight, vij kFor M in kth time iterationij k's
Direction vector, tuFor unit time quantum, vij k+1For the direction vector of+1 iteration of kth, ρ and ξ is Dynamic gene, MPkWith
MGkIt is respectively in kth time iteration the global port data vector of k iteration before current iteration local optimum port data vector sum;
Mij kIt is the first storage matrix, Mij k={ delayij k,LRij k,FLOWij k};
When determining that user gradation optimization priority scheduling evaluation function is not satisfied, or current iteration number of times is less than or equal to maximum
During iterations, it is determined that the flow imbalance of present port, wherein, user gradation optimization priority scheduling evaluation function is not expired
Foot is that the value of user gradation optimization priority scheduling evaluation function is optimum less than or equal to the user gradation of each port in iteration first
Change the meansigma methods of the value of priority scheduling evaluation function.
5. according to the method described in any one of claim 1-4, it is characterised in that the flow obtaining each port at described vCPE is born
Before carrying data, also include:
The terminal of fixed network domestic consumer initiates DHCP request message;Fixing end sets up Vxlan by PG, Vxlan gateway and vCPE
Tunnel, to complete the access of Vxlan double layer network;The DHCP request message of user USER_PC passes through Vxlan double layer network transparent transmission
To vCPE;VCPE supports as DHCP Server to the terminal distribution IP address of lower extension, and supports as IPoE/PPPoE
Client processes and sends to vBRAS through fire wall etc. and carries out dial-up access certification;NAT carries out turning of private network and public network address
Change;NAT is by gateway accessing and accesses Internet;PG passes through the Vxlan tunnel between Vxlan gateway and the vCPE set up
Road, is linked into addressable internal enterprise resources by Vxlan double layer network;The terminal of fixed network domestic consumer is by addressable enterprise
The PG of internal resource accesses the internal office work system of internal enterprise resources such as enterprises and institutions;
Or, mobile terminal device initiates to access the message request of internal enterprise resources, and accesses telecom operators by base station
3G/4G network;Mobile terminal device passes through 3G/4G network insertion VPN server;Mobile terminal device passes through VPN
Server is linked into vBRAS and is authenticated accessing internal enterprise resources.
6. a flow scheduling system based on end office's cloud, it is characterised in that including:
VCPE and each port;
Described vCPE, for obtaining the flow load data of each port;The flow load data of each port that will get, storage
In the first storage matrix;The flow load data of each port in the first storage matrix are optimized lexical analysis, with really
The Optimized Operation strategy of fixed each port;Optimized Operation strategy according to each port, is scheduling the flow of each port.
System the most according to claim 6, it is characterised in that each port, specifically for: every preset time actively to institute
State vCPE and report the flow load data of acquisition;
Or, described vCPE, specifically for: periodically inquire the flow load data that each port obtains.
System the most according to claim 6, it is characterised in that described vCPE, including:
Determine module, for according to user gradation optimization priority scheduling evaluation function, determine that flow, time delay and packet loss lose
The port adjusted;
Adjusting module, for carrying out flow optimization and adjustment to the port of imbalance.
System the most according to claim 8, it is characterised in that described adjusting module, specifically for:
Determine user gradation optimization priority scheduling evaluation function Wherein, xij kFor port whether load traffic definition value and π, ν ∈ (0,1), Qij kAnd Vij kIt is respectively port xij kCongestion Level SPCC based on its priority satisfied
Degree and flow availability, ε andIt is respectively Congestion Level SPCC satisfaction and the Dynamic gene of flow availability;K represents that kth time is repeatedly
Generation, k≤d and k=1,2 ..., d;PRIij k、delayij k、LRij kAnd FLOWij kIt is respectively port priority, port time delay, port
Packet loss and port flow, θ, σ, π and ν are Dynamic gene;
Determine that the first storage matrix after adjustment is Mij k+1=vij k+1*tu, wherein, vij k+1=τij k*vij k+ρ*(MPk-Mij k)+ξ*
(MGk-Mij k),τij kFor autonomic learning pheromone concentration weight, vij kFor M in kth time iterationij k's
Direction vector, tuFor unit time quantum, vij k+1For the direction vector of+1 iteration of kth, ρ and ξ is Dynamic gene, MPkWith
MGkIt is respectively in kth time iteration the global port data vector of k iteration before current iteration local optimum port data vector sum;
Mij kIt is the first storage matrix, Mij k={ delayij k,LRij k,FLOWij k};
When determining that user gradation optimization priority scheduling evaluation function is not satisfied, or current iteration number of times is less than or equal to maximum
During iterations, it is determined that the flow imbalance of present port, wherein, user gradation optimization priority scheduling evaluation function is not expired
Foot is that the value of user gradation optimization priority scheduling evaluation function is optimum less than or equal to the user gradation of each port in iteration first
Change the meansigma methods of the value of priority scheduling evaluation function.
10. according to the system described in any one of claim 6-9, it is characterised in that also include:
The terminal of fixed network domestic consumer and mobile terminal device;
The terminal of fixed network domestic consumer, before obtain the flow load data of each port at described vCPE, initiating DHCP please
Seek message, so that fixing end sets up Vxlan tunnel by PG, Vxlan gateway and vCPE, to complete connecing of Vxlan double layer network
Enter;The DHCP request message of user USER_PC is pass-through to vCPE by Vxlan double layer network;VCPE supports as DHCP
Server gives the terminal distribution IP address of lower extension, and supports to process concurrent as IPoE/PPPoE Client through fire wall etc.
Deliver to vBRAS and carry out dial-up access certification;NAT carries out the conversion of private network and public network address;NAT is by gateway accessing and accesses
Internet;PG passes through the Vxlan tunnel between Vxlan gateway and the vCPE set up, and is linked into by Vxlan double layer network
May have access to internal enterprise resources;The terminal of fixed network domestic consumer accesses enterprises by the PG of addressable internal enterprise resources and provides
The internal office work system of source such as enterprises and institutions;
Mobile terminal device, before obtain the flow load data of each port at described vCPE, initiates to access enterprises
The message request of resource, and the 3G/4G network of telecom operators is accessed by base station;Mobile terminal device passes through 3G/4G network
Access VPN server;Mobile terminal device is linked into vBRAS by VPN server and is authenticated accessing enterprises money
Source.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610819204.7A CN106302221B (en) | 2016-09-12 | 2016-09-12 | Traffic scheduling method and system based on end office's cloud |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610819204.7A CN106302221B (en) | 2016-09-12 | 2016-09-12 | Traffic scheduling method and system based on end office's cloud |
Publications (2)
Publication Number | Publication Date |
---|---|
CN106302221A true CN106302221A (en) | 2017-01-04 |
CN106302221B CN106302221B (en) | 2019-09-10 |
Family
ID=57710173
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610819204.7A Active CN106302221B (en) | 2016-09-12 | 2016-09-12 | Traffic scheduling method and system based on end office's cloud |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106302221B (en) |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106656864A (en) * | 2017-02-21 | 2017-05-10 | 中国联合网络通信集团有限公司 | Message forwarding scheduling method and apparatus |
CN106998303A (en) * | 2017-03-24 | 2017-08-01 | 中国联合网络通信集团有限公司 | The load-balancing method and SiteServer LBS of routing node |
CN108924057A (en) * | 2018-08-02 | 2018-11-30 | 浙江口碑网络技术有限公司 | The port flow intelligence control system of system on a kind of cloud |
CN110768889A (en) * | 2019-10-21 | 2020-02-07 | 烽火通信科技股份有限公司 | VXLAN tunnel construction method and system |
CN110990140A (en) * | 2019-12-10 | 2020-04-10 | 电子科技大学 | Method for scheduling distributed machine learning flow in photoelectric switching network |
CN114039917A (en) * | 2021-11-26 | 2022-02-11 | 中国电信集团系统集成有限责任公司 | Network traffic scheduling comprehensive judgment method and system |
CN114338470A (en) * | 2021-12-21 | 2022-04-12 | 中电信数智科技有限公司 | Port network traffic scheduling judgment method and system |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101827017A (en) * | 2010-02-04 | 2010-09-08 | 蓝盾信息安全技术股份有限公司 | Method and system for dynamically distributing network traffic |
CN102130824A (en) * | 2010-10-30 | 2011-07-20 | 华为技术有限公司 | Joint optimization method, device and system |
CN102984080A (en) * | 2012-12-31 | 2013-03-20 | 无锡城市云计算中心有限公司 | Load balance method used for cloud computation system |
CN104821922A (en) * | 2015-02-12 | 2015-08-05 | 杭州华三通信技术有限公司 | Flow sharing method and equipment |
CN105049262A (en) * | 2015-08-24 | 2015-11-11 | 中国联合网络通信集团有限公司 | Method and device for adjusting flow loads of ports |
-
2016
- 2016-09-12 CN CN201610819204.7A patent/CN106302221B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101827017A (en) * | 2010-02-04 | 2010-09-08 | 蓝盾信息安全技术股份有限公司 | Method and system for dynamically distributing network traffic |
CN102130824A (en) * | 2010-10-30 | 2011-07-20 | 华为技术有限公司 | Joint optimization method, device and system |
CN102984080A (en) * | 2012-12-31 | 2013-03-20 | 无锡城市云计算中心有限公司 | Load balance method used for cloud computation system |
CN104821922A (en) * | 2015-02-12 | 2015-08-05 | 杭州华三通信技术有限公司 | Flow sharing method and equipment |
CN105049262A (en) * | 2015-08-24 | 2015-11-11 | 中国联合网络通信集团有限公司 | Method and device for adjusting flow loads of ports |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106656864A (en) * | 2017-02-21 | 2017-05-10 | 中国联合网络通信集团有限公司 | Message forwarding scheduling method and apparatus |
CN106998303A (en) * | 2017-03-24 | 2017-08-01 | 中国联合网络通信集团有限公司 | The load-balancing method and SiteServer LBS of routing node |
CN108924057A (en) * | 2018-08-02 | 2018-11-30 | 浙江口碑网络技术有限公司 | The port flow intelligence control system of system on a kind of cloud |
CN108924057B (en) * | 2018-08-02 | 2022-06-03 | 浙江口碑网络技术有限公司 | Port flow intelligent control system of on-cloud system |
CN110768889A (en) * | 2019-10-21 | 2020-02-07 | 烽火通信科技股份有限公司 | VXLAN tunnel construction method and system |
CN110768889B (en) * | 2019-10-21 | 2021-11-02 | 烽火通信科技股份有限公司 | VXLAN tunnel construction method and system |
CN110990140A (en) * | 2019-12-10 | 2020-04-10 | 电子科技大学 | Method for scheduling distributed machine learning flow in photoelectric switching network |
CN114039917A (en) * | 2021-11-26 | 2022-02-11 | 中国电信集团系统集成有限责任公司 | Network traffic scheduling comprehensive judgment method and system |
CN114338470A (en) * | 2021-12-21 | 2022-04-12 | 中电信数智科技有限公司 | Port network traffic scheduling judgment method and system |
Also Published As
Publication number | Publication date |
---|---|
CN106302221B (en) | 2019-09-10 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106302221A (en) | Traffic scheduling method based on end office's cloud and system | |
Naranjo et al. | FOCAN: A Fog-supported smart city network architecture for management of applications in the Internet of Everything environments | |
Alam et al. | Autonomic computation offloading in mobile edge for IoT applications | |
Wang et al. | Traffic and computation co-offloading with reinforcement learning in fog computing for industrial applications | |
Kleinrock | Internet congestion control using the power metric: Keep the pipe just full, but no fuller | |
CN105897887B (en) | Remote sensing satellite big data processing system and method based on cloud computing platform | |
Zheng et al. | 5G network-oriented hierarchical distributed cloud computing system resource optimization scheduling and allocation | |
Gonçalves et al. | Dynamic network slicing in fog computing for mobile users in mobfogsim | |
Khan et al. | A cache‐based approach toward improved scheduling in fog computing | |
CN104539744A (en) | Two-stage media edge cloud scheduling method and two-stage media edge cloud scheduling device | |
Zhou et al. | Knowledge transfer and reuse: A case study of AI-enabled resource management in RAN slicing | |
Zhao et al. | Queue-based and learning-based dynamic resources allocation for virtual streaming media server cluster of multi-version VoD system | |
Robles-Enciso et al. | A multi-layer guided reinforcement learning-based tasks offloading in edge computing | |
Jian et al. | A high-efficiency learning model for virtual machine placement in mobile edge computing | |
Ebneyousef et al. | A taxonomy of load balancing algorithms and approaches in fog computing: a survey | |
Javadpour et al. | Mapping and embedding infrastructure resource management in software defined networks | |
Kiran et al. | Optimising resource allocation for virtual network functions in SDN/NFV‐enabled MEC networks | |
Jin et al. | A real-time multimedia streaming transmission control mechanism based on edge cloud computing and opportunistic approximation optimization | |
do Prado et al. | Mobile Edge Computing for Content Distribution and Mobility Support in Smart Cities | |
Ma et al. | A novel approach to cost-efficient scheduling of multi-workflows in the edge computing environment with the proximity constraint | |
Kambalimath et al. | Cost optimization based resource allocation scheme for vehicular cloud networks | |
Kang et al. | Joint task offloading and resource allocation strategy for DiffServ in vehicular cloud system | |
Li et al. | Energy–latency tradeoffs edge server selection and DQN-based resource allocation schemes in MEC | |
Ma et al. | Virtual machine migration strategy based on markov decision and greedy algorithm in edge computing environment | |
Nazari et al. | IETIF: Intelligent Energy‐Aware Task Scheduling Technique in IoT/Fog Networks |
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 |