WO2017133587A1 - 用于通信网络中的流量工程的方法和控制器 - Google Patents
用于通信网络中的流量工程的方法和控制器 Download PDFInfo
- Publication number
- WO2017133587A1 WO2017133587A1 PCT/CN2017/072446 CN2017072446W WO2017133587A1 WO 2017133587 A1 WO2017133587 A1 WO 2017133587A1 CN 2017072446 W CN2017072446 W CN 2017072446W WO 2017133587 A1 WO2017133587 A1 WO 2017133587A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- denotes
- link
- traffic
- node
- communication network
- 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.)
- Ceased
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L43/00—Arrangements for monitoring or testing data switching networks
- H04L43/02—Capturing of monitoring data
- H04L43/022—Capturing of monitoring data by sampling
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/08—Configuration management of networks or network elements
- H04L41/0803—Configuration setting
- H04L41/0823—Configuration setting characterised by the purposes of a change of settings, e.g. optimising configuration for enhancing reliability
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/08—Configuration management of networks or network elements
- H04L41/0895—Configuration of virtualised networks or elements, e.g. virtualised network function or OpenFlow elements
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/14—Network analysis or design
- H04L41/142—Network analysis or design using statistical or mathematical methods
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/16—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks using machine learning or artificial intelligence
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/40—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks using virtualisation of network functions or resources, e.g. SDN or NFV entities
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L43/00—Arrangements for monitoring or testing data switching networks
- H04L43/02—Capturing of monitoring data
- H04L43/022—Capturing of monitoring data by sampling
- H04L43/024—Capturing of monitoring data by sampling by adaptive sampling
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L43/00—Arrangements for monitoring or testing data switching networks
- H04L43/02—Capturing of monitoring data
- H04L43/026—Capturing of monitoring data using flow identification
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L43/00—Arrangements for monitoring or testing data switching networks
- H04L43/08—Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters
- H04L43/0876—Network utilisation, e.g. volume of load or congestion level
- H04L43/0882—Utilisation of link capacity
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L43/00—Arrangements for monitoring or testing data switching networks
- H04L43/08—Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters
- H04L43/0876—Network utilisation, e.g. volume of load or congestion level
- H04L43/0888—Throughput
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L43/00—Arrangements for monitoring or testing data switching networks
- H04L43/12—Network monitoring probes
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L43/00—Arrangements for monitoring or testing data switching networks
- H04L43/20—Arrangements for monitoring or testing data switching networks the monitoring system or the monitored elements being virtualised, abstracted or software-defined entities, e.g. SDN or NFV
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L45/00—Routing or path finding of packets in data switching networks
- H04L45/02—Topology update or discovery
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L45/00—Routing or path finding of packets in data switching networks
- H04L45/12—Shortest path evaluation
- H04L45/123—Evaluation of link metrics
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L45/00—Routing or path finding of packets in data switching networks
- H04L45/42—Centralised routing
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L45/00—Routing or path finding of packets in data switching networks
- H04L45/64—Routing or path finding of packets in data switching networks using an overlay routing layer
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L45/00—Routing or path finding of packets in data switching networks
- H04L45/70—Routing based on monitoring results
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L47/00—Traffic control in data switching networks
-
- 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
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/12—Discovery or management of network topologies
Definitions
- the present application relates to the field of communications, and in particular, to a method and controller for traffic engineering in a communication network.
- Traffic Engineering is a method to control network traffic to achieve certain goals of network integrity, such as optimizing network utilization, equalizing network utilization, reducing network congestion, and reducing network load. technology.
- TE's most common implementation method is to first estimate (or count) all source and destination connection requirements in the entire network, and further utilize various routing mechanisms (or network protocols), according to The TE target is optimized.
- routing mechanisms or network protocols
- Distance Vector routing protocols such as Routing Information Protocol (RIP), Border Gateway Protocol (BGP), and Link State routing protocols, for example, Open Shortest Path First Protocol (Open Shortest Path First, OSPF).
- RIP Routing Information Protocol
- BGP Border Gateway Protocol
- Link State routing protocols for example, Open Shortest Path First Protocol (Open Shortest Path First, OSPF).
- the computational complexity of existing routing protocols can be large (especially for large-scale networks), such as distance vector routing using Bellman-Ford algorithm, with O(
- the method of calculating the best path is to set the short-term link cost to the cost of the distance from the destination, but the TE also needs to be performed multiple times.
- These algorithms are repeatedly run to update the short-term link cost to achieve some of the specific control objectives of the entire network as described above, which shows that the complexity is very large.
- the network-wide traffic demand required for network traffic engineering is estimated based on network state data. If the algorithm running TE is slow, even if the network state data sampling frequency is high, as the network state changes, the TE uses Effective network status information is actually limited (sparse). However, existing TE technologies are based on the assumption that network state data is sufficient and constant. This kind of situation with only sparse network status data, according to the original basis It is difficult to implement TE control in the algorithm of the whole network.
- SDN Software Defined Network
- SDN controller has strong controllability to the network (because the distributed network is different from the previous distributed network, the SDN adopts the architecture of the centralized controller), so the performance of the TE on the SDN is better than the existing network.
- the prior art also requires a full network state.
- the structure of the SDN is centralized (so that the performance of the TE is better than the existing network), and the network traffic demand needs to be estimated before the TE control.
- the relevant information required for the entire network TE is to be sampled and fed back to the SDN controller, the required measurement data feedback overhead will be large.
- the validity of these information is generally lower than the operation time of TE (because, as mentioned above, the high computational complexity of the SDN controller is large due to the large amount of computation required, and it is normal for the network scene to change before the operation result is obtained) .
- the existing SDN technology requires a complete network message (so it is generally intensively sampled), and it is necessary to assume that these network messages remain invariant (ie, constant) during the TE Control operation. It is difficult to implement TE control according to the original network-based algorithm.
- the feedback data can be achieved by the TE optimization problem implemented by the low complexity algorithm, which is equivalent to the problem that the expected effect of TE in the full network environment is the most important.
- Embodiments of the present invention provide a method and a controller for traffic engineering in a communication network, which can implement TE under limited sampling.
- a method for traffic engineering in a communication network comprising a controller and a plurality of nodes, the plurality of nodes comprising a flexible node and a non-flexible node, the flexible node being directly connected to the controller Connection, the non-flexible node is indirectly connected to the controller, and the method includes:
- the controller determines a time frequency of the at least one sampling node and the traffic sample from the flexible node
- the controller notifies the at least one sampling node to perform traffic sampling at the time frequency, and receives traffic sampling data of all sampling nodes of each sampling;
- the controller determines, according to the flow sampling data of each sampling and the total traffic information of the link in the communication network that the link utilization is higher than the first preset value or lower than the second preset value, determining the required traffic engineering Parameter information;
- the controller performs traffic engineering control based on the parameter information.
- the flow sampling is performed by the determined at least one sampling node at a time frequency, the controller notifying the at least one sampling node to sample data at a time frequency, and receiving the flow sampling data of all the sampling nodes of each sampling; and according to the flow Data is used to determine the parameter information required for traffic engineering, and finally flow engineering control is performed based on the parameter information. Efficient high performance TE control with limited sampling is achieved.
- the traffic sampling data includes traffic flows in the communication network that pass through or originate from each of the sampling nodes to each of the destination nodes in the communication network.
- the controller may adopt a node according to an existing notification, for example, a method that can be delivered by a flow table, and the like.
- the embodiment of the present invention does not limit this.
- the controller can obtain the sampling data sent by the node according to the existing manner, which is not limited by the embodiment of the present invention.
- the purpose of traffic engineering control is to achieve some TE goals. For example: reduce network congestion, optimize network utilization, balance network load, and more.
- the controller can achieve partial TE targets by reducing the maximum utilization of links in the communication network.
- the traffic sampling data includes traffic flows in the communication network that pass through or start from each sampling node to each of the destination nodes in the communication network.
- the controller determines a time frequency of the at least one sampling node and the traffic sampling from the flexible node, including:
- the controller determines a time frequency of the at least one sampling node and the traffic sample according to the estimated total traffic of the traffic in the communication network.
- the controller determines, according to the estimated total traffic of the service in the communication network, a time frequency of the at least one sampling node and the traffic sample, including :
- the controller determines the number of the at least one sampling node and the time frequency according to the following formula:
- V*T s ⁇ 12 log(
- V*T s ⁇ 22
- represents the number of sampling nodes
- T s represents the sampling frequency
- V represents the estimated total traffic
- C represents a set of flexible nodes in the communication network
- represents a flexible node in the communication network.
- the number of ⁇ 11 , ⁇ 12 , ⁇ 21 , ⁇ 22 is a preset constant parameter, and 0 ⁇ 11 ⁇ 1, 0 ⁇ log( ⁇ 21
- the controller After determining the number of the sampling nodes, the controller randomly selects
- the probability of randomness is not limited, for example, it may be an equal probability, or may be a ratio of the number of times of access of the node as a probability, or may be a degree by node. The ratio is taken as a decision of the probability, which is not limited by the embodiment of the present invention.
- the controller determines, for each destination node d, the number of the at least one sampling node and the time frequency according to the following formula:
- V d *T s d ⁇ 12 d log(
- the controller After determining the number of sampling nodes, the controller selects
- each selected sampling node is selected from independent paths in the DAG.
- node is selected from the common parts of the DAGs of all destination nodes d.
- each purpose The DAG diagram of the node d may be the same DAG diagram or a different DAG diagram, and the embodiment of the present invention is not limited thereto.
- the parameter information required by the traffic engineering includes injecting service traffic of each flexible node to each destination node in the communication network.
- the link utilization in the communication network is higher than the uncontrollable traffic in the link of the preset value.
- the controller determines, according to the flow sampling data of each sampling and the total traffic information of the link in the communication network that the link utilization is higher than the first preset value or lower than the second preset value, determining the required traffic engineering Parameter information, including parameter information required to determine traffic engineering according to the following formula (1):
- E represents all sets of links in the communication network
- N represents a set of all nodes in the communication network
- C represents a set of all flexible nodes in the communication network
- vec ( ⁇ I) means that all elements ⁇ I (for all u) are composed of vectors.
- e represents a link in the communication network whose link utilization is higher than the first preset value or a link lower than the second preset value
- f(e) represents the total traffic of the link e
- g(e ) indicates the amount of uncontrollable traffic in link e
- W wd represents the traffic flow in the communication network that passes through or starts from the sampling node w to the node d in the communication network
- I ud represents the injected service traffic from the flexible node u to the destination node d
- a packet is sent from node u to destination d, the total share of link e is denoted as ⁇ e (u,d), the total share of sampled node w is expressed as ⁇ w (u,d), and when w is u
- the outgoing package is expressed as u ⁇ w.
- the parameter information required to determine the flow engineering by solving the l 0 norm minimization can also be determined by other alternative methods, for example, by solving the l 1 The norm is minimized, the weighted l 1 norm is minimized, or the parameter information required for traffic engineering is determined by solving the l p norm minimization, where 0 ⁇ p ⁇ 1. The details are described below separately.
- the parameter information required by the traffic engineering includes injecting service traffic of each flexible node to each destination node in the communication network.
- the link utilization in the communication network is higher than the uncontrollable traffic in the link of the preset value.
- the controller determines, according to the flow sampling data of each sampling and the total traffic information of the link in the communication network that the link utilization is higher than the first preset value or lower than the second preset value, determining the required traffic engineering
- the parameter information includes determining parameter information required for traffic engineering according to any one of the following formulas (2), (3), and (4):
- E represents all sets of links in the communication network
- N represents a set of all flexible nodes in the communication network
- C represents a set of all flexible nodes in the communication network
- vec ( ⁇ I) means that all elements ⁇ I (for all u) are composed of vectors.
- e represents a link in the communication network whose link utilization is higher than the first preset value or a link lower than the second preset value
- f(e) represents the total traffic of the link e
- g(e ) indicates the amount of uncontrollable traffic in link e
- W wd represents the traffic flow in the communication network that passes through or starts from the sampling node w to the node d in the communication network
- I ud represents the injected service traffic from the flexible node u to the destination node d
- a packet is sent from node u to destination d, the total share of link e is denoted as ⁇ e (u,d), the total share of sampled node w is expressed as ⁇ w (u,d), and when w is u
- the outgoing package is expressed as u ⁇ w;
- the controller root Determining parameters required for traffic engineering according to the sampled data sample data and the total traffic information of the link in the communication network that is higher than the first preset value or lower than the second preset value.
- Information including formula (2) and formula (3) according to the convex optimization method (which may also include the shaving algorithm), and solving the formula by the continuous convex optimization approximation (which may include continuous use of the shaving algorithm) Any one of (4) determines the parameter information required for the traffic engineering.
- the controller performs traffic engineering control according to the parameter information, including: the controller minimizing the communication network according to the parameter information. The cost of all the paths.
- the controller minimizes a cost of all paths in the communication network according to the parameter information, including the controller according to the following formula ( 5) Network cost of the communication network:
- E denotes a set of all links
- C denotes a set of flexible nodes
- u denotes a flexible node
- d denotes a destination node
- N denotes a set of all nodes
- I ud denotes a flexible node u to a destination node d injects traffic
- P ud represents a set of all feasible paths from u to d
- an example of a feasible path is in a conventional method, such as in OSPF (Open Shortest Path First Protocol), if u' is a point in the path and is a legacy node (non-flexible node), u' in path P
- the next node is the next node in the shortest path to d
- another example of a feasible path may be a network administrator-defined set of paths.
- P ⁇ P ud indicates that P represents a feasible path in P ud .
- x(P) represents the traffic of the path P.
- ⁇ (e) represents the utilization of the controllable bandwidth of link e, ie
- c(e) represents the total bandwidth of link e
- g(e) represents the traffic of uncontrollable flow of link e
- a function constructed by the utilization of the controllable bandwidth of all links e that is, setting the utilization rate of the controllable bandwidth of all links e Will give a real number representing the cost, for example Cost is the maximum utilization of all links; for example The cost is the cost of the path P The sum of weighted by ⁇ P , among them Can be That is, the cost of link e on path P The sum of weighted by ⁇ e , among them It may be the link utilization rate ⁇ (e), and the formula (5) means minimizing the (weighted) most congested place in the communication network; It may also be the reciprocal of the link utilization, i.e., 1/ ⁇ (e), and equation (5) indicates that the (unweighted) most unbalanced or least utilized portion of the communication network is minimized.
- the controller minimizes a cost of all paths in the communication network according to the parameter information, including the controller according to the controller
- the following formula (6) minimizes the maximum cost of all paths in the communication network:
- e denotes the link
- E denotes a set of all links
- C denotes a set of flexible nodes
- u denotes a flexible node
- d denotes a destination node
- N denotes a set of all nodes
- I ud denotes a flexible node u to a destination node d injects traffic
- P ud represents a set of all feasible paths from u to d
- P ⁇ ud indicates that P represents a feasible path in Pud
- x(P) represents the traffic of the path P.
- the controller minimizes a cost of all paths in the communication network according to the parameter information, including the controller according to a formula (5-11) Minimizing network costs in the communication network:
- e denotes the link
- E denotes a set of all links
- C denotes a set of flexible nodes
- u denotes a flexible node
- d denotes a destination node
- N denotes a set of all nodes
- I ud denotes a flexible node u to a destination node d injects traffic
- P ud denotes a set of all feasible paths from u to d
- P ⁇ P ud denotes P denotes a feasible path in Pud
- x(P) denotes traffic of path P
- ⁇ denotes the largest network Throughput ratio
- ⁇ represents the maximum utilization related parameter of the link
- h e ( ⁇ ) is used to represent the relationship function between the weight parameter and ⁇ , wherein the weight parameters include ⁇ e and ⁇ P , ⁇ e represents the weight parameter of link e, the weight parameter of
- h e ( ⁇ ) may be any one of the following:
- weight parameter ⁇ P >0 of path P can be multiplied or exponentially Weighted in The weight parameter ⁇ e >0 of link e can also be weighted by a product method or by an exponential method.
- the formula (5-11) can be transformed into the following formulas (5-12), (5-13), (5-14), and (5-15).
- the controller root Minimizing the cost of all paths in the communication network in accordance with the parameter information, including the controller minimizing network costs in the communication network according to equation (5-21):
- e denotes the link
- E denotes a set of all links
- C denotes a set of flexible nodes
- u denotes a flexible node
- d denotes a destination node
- N denotes a set of all nodes
- I ud denotes a flexible node u to a destination node d injects traffic
- P ud denotes a set of all feasible paths from u to d
- P ⁇ P ud denotes P denotes a feasible path in Pud
- x(P) denotes traffic of path P
- ⁇ denotes the largest network Throughput ratio
- ⁇ represents the equilibrium-related parameters of the link
- h e ( ⁇ ) is used to represent the relationship function between the weight parameters and ⁇ , where the weight parameters include ⁇ e and ⁇ P , ⁇ e Indicates the weight parameter of link e, the weight parameter
- weighting parameters ⁇ e and ⁇ P , h e ( ⁇ ) may be any one of the following:
- the construction represents the network cost
- different weights are set, and the corresponding h e ( ⁇ ) has different forms.
- the weight parameter ⁇ P >0 of path P can be multiplied or exponentially Weighted in
- the weight parameter ⁇ e >0 of link e can also be weighted by the product method or by the exponential method.
- the formula (5-21) can be transformed into the following formulas (5-22), (5-23) (5-24), and (5-25).
- the controller minimizes a cost of all paths in the communication network according to the parameter information, including the controller according to a formula (6-11) Minimizing network costs in the communication network:
- e denotes the link
- E denotes a set of all links
- C denotes a set of flexible nodes
- u denotes a flexible node
- d denotes a destination node
- N denotes a set of all nodes
- I ud denotes a flexible node u to a destination node d injects traffic
- P ud denotes a set of all feasible paths from u to d
- P ⁇ ud denotes P denotes a feasible path in P ud
- x (P) denotes traffic of path P
- ⁇ represents the maximum network throughput related parameter
- k e ( ⁇ ) is used to indicate the correspondence between the weight parameter of link e and ⁇
- ⁇ (e) represents the utilization of the controllable bandwidth of link e
- c(e) represents the total bandwidth of link e
- g(e) represents
- the controller root (6-11) minimizes network cost in the communication network, including:
- the controller takes the k e ( ⁇ ) as a constant and uses a low complexity maximum concurrent fractional flow algorithm to solve the equation (6-11) to minimize network costs in the communication network.
- the controller can minimize the complexity in the communication network by considering the k e ( ⁇ ) in (6-11) as a constant (ie, the current iteration parameter) with a low complexity maximum concurrent fractional flow algorithm.
- Network cost ie, the current iteration parameter
- the short-term link cost in the maximum concurrent fractional flow algorithm will be related to the current iteration parameter, for example, the short-term path cost is proportional or positively correlated with the link utilization.
- the controller minimizes the cost of all paths in the communication network based on the parameter information, including the controller minimizing the maximum cost of all paths in the communication network according to formula (6-12):
- e denotes the link
- E denotes a set of all links
- C denotes a set of flexible nodes
- u denotes a flexible node
- d denotes a destination node
- N denotes a set of all nodes
- I ud denotes a flexible node u to a destination node d injects traffic
- P ud denotes a set of all feasible paths from u to d
- P ⁇ ud denotes P denotes a feasible path in P ud
- x (P) denotes traffic of path P
- ⁇ represents the maximum network throughput related parameter
- k e ( ⁇ ) is used to represent the correspondence between the weight parameter of link e and ⁇ , where ⁇ e represents the link Weight parameter, ⁇ (e) represents the utilization of the controllable bandwidth of link e, c(e) represents the total bandwidth of link
- Equation (6-12) minimizes network costs in the communication network, including:
- the control takes k e ( ⁇ ) as a constant and solves the formula (6-12) with a low complexity maximum concurrent fractional flow algorithm to minimize the maximum cost of all paths in the communication network.
- the controller can minimize the complexity of the communication network by using k e ( ⁇ ) in (6-12) as a constant (ie, the current iteration parameter) with a low complexity maximum concurrent fractional flow algorithm.
- Network cost k e ( ⁇ ) in (6-12) as a constant (ie, the current iteration parameter) with a low complexity maximum concurrent fractional flow algorithm.
- Network cost k e ( ⁇ ) in (6-12) as a constant (ie, the current iteration parameter) with a low complexity maximum concurrent fractional flow algorithm.
- Network cost Among them, the short-term link cost in the maximum concurrent fractional flow algorithm will be related to the current iteration parameter, for example, the short-term path cost is proportional or positively correlated with the link utilization.
- the controller minimizes a cost of all paths in the communication network according to the parameter information, including the controller according to a formula (6-21) Minimize network costs in the communication network:
- e denotes the link
- E denotes a set of all links
- C denotes a set of flexible nodes
- u denotes a flexible node
- d denotes a destination node
- N denotes a set of all nodes
- I ud denotes a flexible node u to a destination node d implantation traffic
- P ud from u represents the set of all possible paths to d
- P ⁇ P ud represents P represents a feasible path
- P UD is
- x (P) denotes the flow path P
- ⁇ represents the balance network Network throughput related parameters
- k e ( ⁇ ) is used to indicate the correspondence between the weight parameter of link e and ⁇
- the weight parameters include ⁇ e and ⁇ P
- ⁇ e represents the link weight parameter
- ⁇ P represents the path weight parameter
- ⁇ (e) represents the utilization of the controllable bandwidth
- the controller formula (6-21) minimizes network cost in the communication network, including:
- the controller takes the k e ( ⁇ ) as a constant and minimizes the network cost in the communication network using a low complexity Maximum Current Concurrent Flow solution (6-21).
- the controller can treat k e ( ⁇ ) in (6-21) as a constant (that is, the current iteration parameter), that is, adopt any one of them, and the minimum concurrent maximum fractional flow algorithm with low complexity is minimal.
- the network cost in the communication network Among them, the short-term link cost in the maximum concurrent fractional flow algorithm will be related to the current iteration parameter, for example, the short-term path cost is proportional or positively correlated with the link utilization.
- the controller minimizes a cost of all paths in the communication network according to the parameter information, including the controller according to a formula (6-22) Minimize the maximum cost of all paths in the communication network:
- e denotes the link
- E denotes a set of all links
- C denotes a set of flexible nodes
- u denotes a flexible node
- d denotes a destination node
- N denotes a set of all nodes
- I ud denotes a flexible node u to a destination node d implantation traffic
- P ud from u represents the set of all possible paths to d
- P ⁇ P ud represents P represents a feasible path P UD is
- x (P) denotes the flow path P
- ⁇ represents the balance network Network throughput related parameters
- k e ( ⁇ ) is used to represent the correspondence between the weight parameter of link e and ⁇ , where the weight parameters include ⁇ e and ⁇ P , and ⁇ e represents the link weight parameter
- ⁇ (e) represents the utilization of the controllable bandwidth of the link e
- the controller minimizes network cost in the communication network according to formula (6-22), including:
- the controller takes the k e ( ⁇ ) as a constant and solves the formula (6-22) with an improved maximum concurrent fractional stream algorithm to minimize network costs in the communication network.
- the controller considers k e ( ⁇ ) in (6-22) as a constant (ie, the current iteration parameter), and uses a revised maximum concurrent fractional flow algorithm (ie, the original maximum concurrent fractional flow algorithm).
- the short-term link cost in the middle is considered positive, above which a new short-term link cost associated with this current iteration parameter is set and treated as a negative value) to minimize the network cost in the communication network .
- the present invention can implement fast TE control, and can implement existing TE control problems that are considered to be high complexity by different prior art technologies with low complexity algorithms.
- the embodiment of the present invention proposes and solves the more general TE control problem, so that the real-time TE can be implemented.
- a controller in a communication network is provided, the controller being capable of implementing any of the first aspect and its implementation, wherein the operations and/or functions of the respective modules in the controller are respectively used
- the corresponding method features in the first aspect of the implementation and the implementation manner thereof are not described herein for brevity.
- a controller in a communication network comprising a memory and a processor storing instructions, wherein the processor executes the instructions to perform any of the first aspect and various implementations thereof A method for traffic engineering in a communication network.
- a processing apparatus is provided that is applied to a communication system.
- the processing device can be one or more processors or chips.
- the processing device may also be a physical device or a virtual device in a communication system.
- the processing device is configured to perform the method of any of the above first aspects and various implementations thereof for traffic engineering in a communication network.
- a computer program product comprising: computer program code, when the computer program code is executed by a computing unit, a processing unit or a processor of a communication device, causing the communication device to perform the first Any of the aspects and various implementations thereof for a method of traffic engineering in a communication network.
- a computer readable storage medium storing a program causing a communication device to perform any of the first aspect described above and various implementations thereof for use in a communication network The method of traffic engineering.
- the embodiment of the present invention performs traffic sampling by using the determined at least one sampling node at a time frequency, and the controller notifies the at least one sampling node to sample data at a time frequency, and receives traffic of all sampling nodes of each sampling. Sampling data; and using the data to determine the parameter information required for the traffic engineering according to the flow, and finally performing flow engineering control according to the parameter information.
- TE control under limited sampling is achieved.
- FIG. 1 is a schematic diagram of an SDN network system that can be used in an embodiment of the present invention.
- FIG. 2 is a schematic flow diagram of a method for traffic engineering in a communication network, in accordance with one embodiment of the present invention.
- FIG. 3 is a schematic flow diagram of a method for traffic engineering in a communication network in accordance with another embodiment of the present invention.
- FIG. 4 is a schematic block diagram of a controller device in a communication network in accordance with one embodiment of the present invention.
- FIG. 5 is a schematic block diagram of a controller device in a communication network according to another embodiment of the present invention.
- the examples of the present invention can be applied to various communication networks (especially the network layer of a communication system), such as an IP network, an MPLS-based network, and an ad hoc network (Ad).
- -hoc network ad hoc network
- Information Centric Network Content Distribution Network
- SDN ad hoc network
- GSM Global System of Mobile communication
- CDMA Code Division Multiple Access
- WCDMA Wideband Code Division Multiple Access
- GPRS General Packet Radio Service
- LTE Long Term Evolution
- FDD Frequency Division Duplex
- TDD Time Division Duplex
- UMTS Universal Mobile Telecommunication System
- WiMAX Worldwide Interoperability for Microwave Access
- FIG. 1 is a schematic diagram of an SDN network system that can be used in an embodiment of the present invention.
- the SDN network architecture shown in FIG. 1 includes a controller and nodes 1 to 15, wherein nodes 2, 9, and 14 are directly connected to the controller as flexible nodes; nodes 1, 3, 4, 5, 6, and 7, 8, 10, 11, 12, 13, 15 are indirectly connected to the controller indirectly, which is a non-flexible node.
- the sub-set C is called an SDN-Forwarding Element (SDN-FE; that is, a Flexible Node or a Type 1 node). They are directly connected to an SDN Controller (SDN Controller).
- SDN-FE SDN-Forwarding Element
- SDN-FE SDN-Forwarding Element
- SDN-FE SDN-Forwarding Element
- SDN Controller SDN Controller
- the other nodes are called non-flexible nodes or class 2 nodes, and the set is represented as D. That is, the flexible node set is C, and the non-flexible node set is D.
- the number on the node represents the number of the node
- the set N contains all the nodes, that is, the node 1 to the node 15, and the set C includes ⁇ 2, 9, 14 ⁇ , and the three nodes
- the set D includes ⁇ 1, 3,4,5,6,7,8,10,11,12,13,15 ⁇ .
- Uncontrollable Traffic A flow that travels from a source node to a destination node without passing through a flexible node is called Uncontrollable Traffic.
- the total traffic is represented as f(e) and the uncontrollable portion is represented as g(e).
- T sd represents the total traffic from the source node s in N to the destination node in N.
- W ud represents the total traffic passing or starting with the destination node d in nodes u to N in C.
- I ud represents the total traffic injected by the node u in C to the destination node d in N.
- the communication network includes a controller. And a plurality of nodes, the flexible node and the non-flexible node, the flexible node being directly connected to the controller, the non-flexible node being indirectly connected to the controller, the method being executable by the controller, the method comprising:
- the controller determines, from the flexible node, a time frequency of the at least one sampling node and the traffic sample.
- the controller notifies the at least one sampling node to perform traffic sampling at the time frequency, and receives traffic sampling data of all sampling nodes of each sampling.
- the controller determines, according to the sampled data of each sampling, the total traffic information of the link in the communication network that is higher than the first preset value or lower than the second preset value, and determines the traffic engineering office. Required parameter information.
- the controller performs flow engineering control according to the parameter information.
- the embodiment of the present invention performs traffic sampling by using the determined at least one sampling node at a time frequency, and the controller notifies the at least one sampling node to sample data at a time frequency, and receives traffic sampling data of all sampling nodes of each sampling; According to the traffic, the data is used to determine the parameter information required by the traffic engineering, and finally the traffic engineering control is performed according to the parameter information. TE control under limited sampling is achieved.
- the traffic sampling data includes traffic flows in the communication network that pass through or originate from each of the sampling nodes to each of the destination nodes in the communication network.
- the controller may adopt a node according to the existing notification, for example, a method that can be delivered by using a flow table, and the like, which is not limited by the embodiment of the present invention.
- the controller may also implement the method according to the existing manner. The controller obtains the sampling data sent by the node, which is not limited by the embodiment of the present invention.
- the controller determines a time frequency of the at least one sampling node and the traffic sample according to the estimated total traffic of the service in the communication network.
- the controller determines the number of the at least one sampling node and the time frequency according to the following formula:
- V*T s ⁇ 12 log(
- V*T s ⁇ 22
- represents the number of sampling nodes
- T s represents the sampling frequency
- V represents the estimated total traffic
- C represents a set of flexible nodes in the communication network
- represents a flexible node in the communication network.
- the number of ⁇ 11 , ⁇ 12 , ⁇ 21 , ⁇ 22 is a preset constant parameter, and 0 ⁇ 11 ⁇ 1, 0 ⁇ log( ⁇ 21
- the controller After determining the number of sampling nodes, the controller randomly selects
- the probability of randomness is not limited, for example, it may be an equal probability, or may be a ratio of the number of times of access of the node as a probability, or may be a degree by node. The ratio is taken as a decision of the probability, which is not limited by the embodiment of the present invention.
- the controller determines, for each destination node d, the number of the at least one sampling node and the time frequency according to the following formula:
- V d *T s d ⁇ 12 d log(
- the controller selects
- DAG Directed Acyclic Graph
- each of the selected nodes are selected out of the sample from the DAG independent path, for example selected from a common portion common to all destination nodes d DAG in
- the parameter information required by the traffic engineering includes injecting service traffic of each flexible node to each destination node in the communication network, and link utilization in the communication network is higher than a pre-pre Uncontrollable traffic in a set-valued link,
- the controller determines the parameter information required for traffic engineering according to the following formula (1):
- E represents all sets of links in the communication network
- N represents a set of all nodes in the communication network
- C represents a set of all flexible nodes in the communication network
- vec ( ⁇ I) means that all elements ⁇ I (for all u) are composed of vectors.
- a minimum l 0 norm representing vec( ⁇ I) where e represents a link in the communication network with a link utilization higher than a first preset value or a link lower than a second preset value, f(e Indicates the total traffic of link e, g(e) represents the uncontrollable traffic in link e, and W wd represents the traffic flow in the communication network that passes through or starts from the sampling node w to node d in the communication network.
- I ud represents the injected service traffic from the flexible node u to the destination node d
- a packet is sent from node u to destination d, the total share of link e is denoted as ⁇ e (u,d), the total share of sampled node w is expressed as ⁇ w (u,d), and when w is u
- the outgoing package is expressed as u ⁇ w.
- the parameter information required to determine the flow engineering by solving the l 0 norm minimization can also be determined by other alternative methods, for example, by solving the l 1 The norm is minimized, the weighted l 1 norm is minimized, or the parameter information required for traffic engineering is determined by solving the l p norm minimization, where 0 ⁇ p ⁇ 1. The details are described below separately.
- the controller determines parameter information required for traffic engineering according to any one of the following formulas (2), (3), and (4):
- E represents all sets of links in the communication network
- N represents a set of all flexible nodes in the communication network
- C represents a set of all flexible nodes in the communication network
- vec ( ⁇ I) means that all elements ⁇ I (for all u) are composed of vectors.
- e represents a link in the communication network whose link utilization is higher than the first preset value or a link lower than the second preset value
- f(e) represents the total traffic of the link e
- g(e ) indicates the amount of uncontrollable traffic in link e
- W wd represents the traffic flow in the communication network that passes through or starts from the sampling node w to the node d in the communication network
- I ud represents the injected service traffic from the flexible node u to the destination node d
- Equation (3) represents the solution weighted l 1 norm minimization, ie
- Equation (4) indicates that the l p norm is minimized, or the weighted l p norm is minimized (where 0 ⁇ p ⁇ 1, that is, the formula (2), or the l 1 norm of the formula (3) is changed to l p Norm.
- the purpose of flow engineering control is to achieve some TE goals. For example: reduce network congestion, optimize network utilization, balance network load, and more.
- the controller can achieve partial TE targets by reducing the maximum utilization of links in the communication network.
- the controller minimizes the cost of all paths in the communication network based on the parameter information.
- the controller minimizes the network cost of the communication network according to the following formula (5):
- e denotes the link
- E denotes a set of all links
- C denotes a set of flexible nodes
- u denotes a flexible node
- d denotes a destination node
- N denotes a set of all nodes
- I ud denotes a flexible node u to a destination node d injects traffic
- P ud represents a set of all feasible paths from u to d.
- an example of a feasible path is in a conventional method, such as in Open Shortest Path First (OSPF), if u' is One point in the path is also a traditional node (non-flexible node). The next node in u' of path P is the next node in the shortest path to d.
- Another example of a feasible path can be customized by the network administrator. Path collection.
- P ⁇ P ud indicates that P represents a feasible path in P ud .
- x(P) represents the traffic of the path P. Represents the total traffic of all feasible paths from u to d, ⁇ (e) represents the utilization of the controllable bandwidth of link e, ie Where c(e) represents the total bandwidth of link e, and g(e) represents the traffic of uncontrollable flow of link e, The total flow of controllable flows representing all feasible paths on link e; Represents network costs.
- a function constructed by the utilization of the controllable bandwidth of all links e on the path P that is, setting the utilization rate of the controllable bandwidth of all links e on the path P
- the controller minimizes the maximum cost of all paths in the communication network according to the following formula (6):
- e denotes the link
- E denotes a set of all links
- C denotes a set of flexible nodes
- u denotes a flexible node
- d denotes a destination node
- N denotes a set of all nodes
- I ud denotes a flexible node u to a destination node d injects traffic
- P ud represents a set of all feasible paths from u to d
- P ⁇ ud indicates that P represents a feasible path in Pud
- x(P) represents the traffic of the path P.
- the controller according to the parameter information, according to the solution or approximate solution of the equivalent conversion of formula (5) or (6) and its implementation method, in particular, according to the special example of (5), Such as formula (5-11), formula (5-12), formula (5-13), formula (5-14) and formula (5-15); and formula (5-21), formula (5-22) , formula (5-23), formula (5-24), and formula (5-25); and special example formula (6-11) and formula (6-12) of formula (6); and formula (6-21) And formula (6-22) minimizes the path cost.
- formula (5-11), formula (5-12), formula (5-13), formula (5-14) and formula (5-15) and formula (5-21), formula (5-22) , formula (5-23), formula (5-24), and formula (5-25
- special example formula (6-11) and formula (6-12) of formula (6) and formula (6-21) And formula (6-22) minimizes the path cost.
- the controller is minimum according to the following formula: formula (5-11), formula (5-12), formula (5-13), formula (5-14), and formula (5-15)
- formula (5-11) The network cost in the communication network.
- the controller minimizes the cost of all paths in the communication network according to the parameter information, including the controller minimizing network cost in the communication network according to formula (5-11):
- e denotes the link
- E denotes a set of all links
- C denotes a set of flexible nodes
- u denotes a flexible node
- d denotes a destination node
- N denotes a set of all nodes
- I ud denotes a flexible node u to a destination node d injects traffic
- P ud denotes a set of all feasible paths from u to d
- P ⁇ ud denotes P denotes a feasible path in P ud
- x (P) denotes traffic of path P
- ⁇ represents the maximum network throughput ratio
- ⁇ represents the maximum utilization related parameter of the link
- h e ( ⁇ ) is used to represent the relationship function between the weight parameter and ⁇ , where
- the weight parameters include ⁇ e and ⁇ P , ⁇ e represents the weight parameter of link e, the weight parameter
- h e ( ⁇ ) may be any one of the following:
- weight parameter ⁇ P >0 of path P can be multiplied or exponentially Weighted in The weight parameter ⁇ e >0 of link e can also be weighted by a product method or by an exponential method.
- the formula (5-11) can be transformed into the following formulas (5-12), (5-13), (5-14), and (5-15).
- the controller is minimum according to the following formula: formula (5-21), formula (5-22), formula (5-23), formula (5-24), and formula (5-25)
- formula (5-21) formula (5-22), formula (5-23), formula (5-24), and formula (5-25)
- the controller minimizes the cost of all paths in the communication network according to the parameter information, including the controller minimizing network cost in the communication network according to formula (5-21):
- e denotes the link
- E denotes a set of all links
- C denotes a set of flexible nodes
- u denotes a flexible node
- d denotes a destination node
- N denotes a set of all nodes
- I ud denotes a flexible node u to a destination node d injects traffic
- P ud denotes a set of all feasible paths from u to d
- P ⁇ P ud denotes P denotes a feasible path in Pud
- x(P) denotes traffic of path P
- ⁇ denotes the largest network Throughput ratio
- ⁇ represents the equilibrium-related parameters of the link
- h e ( ⁇ ) is used to represent the relationship function between the weight parameters and ⁇ , where the weight parameters include ⁇ e and ⁇ P , ⁇ e Indicates the weight parameter of link e, the weight parameter
- h e ( ⁇ ) may be any one of the following:
- the construction represents the network cost
- different weights are set, and the corresponding h e ( ⁇ ) has different forms.
- the weight parameter ⁇ P >0 of path P can be multiplied or exponentially Weighted in
- the weight parameter ⁇ e >0 of link e can also be weighted by a product method or by an exponential method.
- the formula (5-21) can be transformed into the following formulas (5-22), (5-23) (5-24), and (5-25).
- the controller minimizes the network cost of all paths in the communication network according to any of the following formulas: Equation (6-11) and Equation (6-12).
- the controller minimizes the cost of all paths in the communication network according to the parameter information, including the controller minimizing the network cost in the communication network according to formula (6-11):
- e denotes the link
- E denotes a set of all links
- C denotes a set of flexible nodes
- u denotes a flexible node
- d denotes a destination node
- N denotes a set of all nodes
- I ud denotes a flexible node u to a destination node d injects traffic
- P ud denotes a set of all feasible paths from u to d
- P ⁇ ud denotes P denotes a feasible path in P ud
- x (P) denotes traffic of path P
- ⁇ represents the maximum network throughput related parameter
- k e ( ⁇ ) is used to indicate the correspondence between the weight parameter of link e and ⁇
- ⁇ (e) represents the utilization of the controllable bandwidth of link e
- c(e) represents the total bandwidth of link e
- g(e) represents
- controller root (6-11) minimizes network costs in the communication network, including:
- the controller takes the k e ( ⁇ ) as a constant and uses a low complexity maximum concurrent fractional flow algorithm to solve the equation (6-11) to minimize network costs in the communication network.
- the controller can minimize the complexity in the communication network by considering the k e ( ⁇ ) in (6-11) as a constant (ie, the current iteration parameter) with a low complexity maximum concurrent fractional flow algorithm.
- Network cost ie, the current iteration parameter
- the short-term link cost in the maximum concurrent fractional flow algorithm will be related to the current iteration parameter, for example, the short-term path cost is proportional or positively correlated with the link utilization.
- the controller minimizes the cost of all paths in the communication network based on the parameter information, including the controller minimizing the maximum cost of all paths in the communication network according to formula (6-12) :
- e denotes the link
- E denotes a set of all links
- C denotes a set of flexible nodes
- u denotes a flexible node
- d denotes a destination node
- N denotes a set of all nodes
- I ud denotes a flexible node u to a destination node d injects traffic
- P ud denotes a set of all feasible paths from u to d
- P ⁇ ud denotes P denotes a feasible path in P ud
- x (P) denotes traffic of path P
- ⁇ represents the maximum network throughput related parameter
- k e ( ⁇ ) is used to represent the correspondence between the weight parameter of link e and ⁇ , where ⁇ e represents the link Weight parameter, ⁇ (e) represents the utilization of the controllable bandwidth of link e, c(e) represents the total bandwidth of link
- the controller has the following formula: Equation (6-12) minimizes network costs in the communication network, including:
- the control takes k e ( ⁇ ) as a constant and solves the formula (6-12) with a low complexity maximum concurrent fractional flow algorithm to minimize the maximum cost of all paths in the communication network.
- the controller may (6-12) is a k e ( ⁇ ) as a constant (i.e., the parameters of the current iteration), the maximum concurrent flow fraction low complexity algorithm for minimizing the communication network Network cost.
- the short-term link cost in the maximum concurrent fractional flow algorithm will be related to the current iteration parameter, for example, the short-term path cost is proportional or positively correlated with the link utilization.
- the controller minimizes the maximum cost of all paths in the communication network according to any of the following formulas: Equation (6-21) and Equation (6-22).
- the controller minimizes the cost of all paths in the communication network according to the parameter information, including the controller minimizing network cost in the communication network according to formula (6-21):
- e denotes the link
- E denotes a set of all links
- C denotes a set of flexible nodes
- u denotes a flexible node
- d denotes a destination node
- N denotes a set of all nodes
- I ud denotes a flexible node u to a destination node d implantation traffic
- P ud from u represents the set of all possible paths to d
- P ⁇ P ud represents P represents a feasible path P UD is
- x (P) denotes the flow path P
- ⁇ represents the balance network Network throughput related parameters
- k e ( ⁇ ) is used to indicate the correspondence between the weight parameter of link e and ⁇
- the weight parameters include ⁇ e and ⁇ P , ⁇ e represents the link weight parameter, ⁇ P represents the path weight parameter, ⁇ (e) represents the utilization of the controllable bandwidth of the link e
- the controller minimizes network cost in the communication network according to formula (6-21), including:
- the controller takes the k e ( ⁇ ) as a constant and solves the formula (6-21) with a modified maximum concurrent fractional stream algorithm to minimize network costs in the communication network.
- the controller can treat k e ( ⁇ ) in (6-21) as a constant (that is, the current iteration parameter), that is, adopt any one of them, and the minimum concurrent maximum fractional flow algorithm with low complexity is minimal.
- the network cost in the communication network Among them, the short-term link cost in the maximum concurrent fractional flow algorithm will be related to the current iteration parameter, for example, the short-term path cost is proportional or positively correlated with the link utilization.
- the controller minimizes the cost of all paths in the communication network based on the parameter information, including the controller minimizing the maximum cost of all paths in the communication network according to formula (6-22) :
- e represents the link
- E represents the set of all the links
- C denotes the set of nodes and flexible
- u represents a flexible node
- d represents the destination node
- N represents the set of all nodes
- I ud represents flexible node to the destination node u d implantation traffic
- P ud from u represents the set of all possible paths to d
- P ⁇ P ud represents P represents a feasible path P UD is
- x (P) denotes the flow path P
- ⁇ represents the balance network Network throughput related parameters
- k e ( ⁇ ) is used to represent the correspondence between the weight parameter of link e and ⁇ , where the weight parameters include ⁇ e and ⁇ P , and ⁇ e represents the link weight parameter
- ⁇ P represents the path weight parameter
- ⁇ (e) represents the utilization of the controllable bandwidth of the link e
- c(e) represents the total bandwidth of the
- the controller minimizes network cost in the communication network according to formula (6-22), including:
- the control takes k e ( ⁇ ) as a constant and solves the formula (6-22) with an improved maximum concurrent fractional stream algorithm to minimize the maximum cost of all paths in the communication network.
- the controller considers k e ( ⁇ ) in (6-22) as a constant (ie, the current iteration parameter), and uses a revised maximum concurrent fractional flow algorithm (ie, the original maximum concurrent fractional flow algorithm).
- the short-term link cost in the middle is considered positive, above which a new short-term link cost associated with this current iteration parameter is set and treated as a negative value) to minimize the network cost in the communication network .
- the present invention can implement fast TE control, and can implement existing TE control problems that are considered to be high complexity by different prior art technologies with low complexity algorithms.
- the embodiment of the present invention proposes and solves the more general TE control problem, so that the real-time TE can be implemented.
- the real-time or non-real-time flow control can be performed according to the method of FIG. 2, which is not limited by the embodiment of the present invention.
- the method of traffic engineering according to the embodiment of the present invention is described by using only an SDN network as an example, but the embodiment of the present invention is not limited to the SDN network. That is to say, the traffic engineering method of the embodiment of the present invention can also be applied to other networks.
- the foregoing method is still applicable, for example, to other networks, because there are no flexible nodes and non-flexible nodes.
- the points are collectively referred to as nodes. Therefore, the process engineering control can be realized by considering the flexible nodes in the embodiment of the present invention as nodes in the network, and such modifications are also within the scope of the present invention.
- the traffic engineering shown in Figure 3 is mainly divided into three major processes: the first process: Measurement/Measurement; here is Partial Measurement.
- the second process Network Traffic Model (for TE input recovery); network traffic model
- the main function of the type is to restore the input parameters; that is, to use the results of the first step to restore the input parameters required for the third step control process. Because the result of the first step is usually not the input parameters required for the control process.
- Control Process Its main function is to perform traffic engineering through the input parameters related to network traffic recovered in the second step, that is, through network-wide route optimization to achieve the TE target.
- 310-340 in FIG. 3 corresponds to the first process described above, and 350 corresponds to the second process, and 360 corresponds to the third process.
- each step can be correspondingly The function module is executed.
- the various functional modules herein are only divided for convenience of description. In practical applications, the functions of the corresponding functional modules are implemented by corresponding physical units.
- each functional module may be a processor, which may be executed by a processor. Implement the function of the function module. The embodiments of the present invention are not limited thereto.
- the method 300 includes:
- This process can be performed by a Topology Identification Module, and the description of the present invention is based on the assumption that the topology is known. Among them, one way to obtain a topology is through a Link Layer Discovery Protocol (LLDP).
- LLDP Link Layer Discovery Protocol
- the embodiments of the present invention are not limited thereto.
- the process may be performed by a sampling frequency and a location module (Sampling Frequency and Location Module).
- the traffic measurement may be performed once to obtain the sum of the traffic in one time unit (the network node is first known from the topology identification module). Location and link), or make a preliminary estimate, such as using historical data, as a preliminary estimate. The details will be explained below.
- the process can be performed by a sampling frequency and a positioning module (Sampling Frequency and Location Module), and the sampling frequency and positioning module obtains the network node position and link according to 310, and then determines the sampling frequency according to the predicted traffic of the network traffic prediction module ( In other words, determine how many time units the sampling time is and the sampling position (in a sampling time interval).
- a sampling frequency and a positioning module Samling Frequency and Location Module
- the SDN controller informs the SDN forwarding device (SDN-FE) of the relevant sampling device of the traffic sampling requirement (the determined sampling ratio).
- SDN-FE SDN forwarding device
- the controller determines the number of the at least one sampling node and the time frequency according to the following formula:
- V*T s ⁇ 12 log(
- V*T s ⁇ 22
- represents the number of sampling nodes
- T s represents the sampling frequency
- V represents the estimated total traffic
- C represents a set of flexible nodes in the communication network
- represents a flexible node in the communication network.
- the number of ⁇ 11 , ⁇ 12 , ⁇ 21 , ⁇ 22 is a preset constant parameter, and 0 ⁇ 11 ⁇ 1, 0 ⁇ log( ⁇ 21
- the controller After determining the number of the sampling nodes, the controller randomly selects
- the probability of randomness is not limited, for example, it may be an equal probability, or may be a ratio of the number of times of access of the node as a probability, or may be a degree by node. The ratio is taken as a decision of the probability, which is not limited by the embodiment of the present invention.
- the controller determines, for each destination node d, the number of the at least one sampling node and the time frequency according to the following formula:
- V d *T s d ⁇ 12 d log(
- the controller selects
- DAG Directed Acyclic Graph
- each of the selected nodes are selected out of the sample from the DAG independent path, for example selected from a common portion common to all destination nodes d DAG in
- This process can be performed by the Sampling Module, which samples some TE-related data for the purpose of Traffic Engineering (TE).
- the present invention needs to sample the passing stream W at the time of each sampling time interval according to the sampling frequency and the time specified by the positioning module and the position of the node.
- the specific adoption process can be performed according to the existing method, and the present invention does not limit the specific sampling method.
- sampling node will sample according to the sampling frequency sampling position.
- the specified sampling method and the specific sampling method and the flow may be performed by using the prior art.
- the method may be performed as follows, but the embodiment of the present invention is not limited thereto.
- the controller matches the flow table entry information with the sampling node information, generates a group table entry (including a sampling ratio to perform a sampling behavior), and directs the flow table entry (including the flow information) and sends it to the sampling node, so that The nodes are sampled according to the frequency of adoption.
- sampling node sends the sampled measurement data to the controller, ie the controller obtains the measurement data.
- the process can be performed by an Input Recovery Module, which infers from the measurement data the value of the factor that needs (or can be) used for flow control (TE). (Each of each operation time is one. The sampling time is in units.)
- Its input is the available network measurement result (ie, the output of the sampling module); in an embodiment, it is the flow of each destination d and each SDN-FE node u in one time unit, (u,d) and the total traffic of each link e totalal, f(e). That is, in a sampling time interval, for each d, W and f(e) is its input.
- Its output is the network state available to the TE; in an embodiment, in one time unit, it is the injection stream for each destination d and each SDN-FE node u, I(u,d) and each The uncontrollable traffic of link e, g(e). That is, in one sampling time interval, for each d, I and g are its outputs.
- the controller can determine the parameter information required for the traffic engineering according to formula (1):
- the parameter information required to determine the flow engineering by solving the l 0 norm minimization can also be determined by other alternative methods, for example, by solving the l 1 The norm is minimized, the weighted l 1 norm is minimized, or the parameter information required for traffic engineering is determined by solving the l p norm minimization, where 0 ⁇ p ⁇ 1. The details are described below separately.
- the controller can determine the parameter information required for the traffic engineering according to any one of the formula (2), the formula (3), and the formula (4).
- This process can be performed by the Network Traffic Control Module (TE Control Module), which aims to achieve some TE objectives. For example: reduce network congestion, optimize network utilization, balance network load, and more.
- the controller can achieve partial TE targets by reducing the maximum utilization of links in the communication network.
- each operation time in the process engineering can be a unit of time or a sampling time interval. Its input is the available network state (ie the output of the input recovery module). Its output is for each flexible node u and destination node d, which gives each path from u to d and its traffic.
- the controller minimizes the cost of all paths in the communication network based on the parameter information.
- controller minimizes the cost of all paths in the communication network according to equation (5) or (6).
- the controller according to a specific example according to (5), such as formula (5-11), formula (5-12), formula (5-13), formula (5-14) And formula (5-15); and formula (5-21), formula (5-22), formula (5-23), formula (5-24), and formula (5-25); and formula (6)
- formula (6-11) and the formula (6-12; and the formula (6-21) and the formula (6-22) are used to minimize the path cost.
- the controller in the embodiment of the present invention may further include other modules.
- the communication module is mainly responsible for communication between the modules, so there may or may not be, and the functions may also be put into other modules. .
- this module will cause the SDN controller to sample the frequency and the flow sampling requirements specified by the positioning module (including the sampling ratio), and the flow table information (including the boot flow table entries) manufactured by the network traffic engineering control module.
- Send to the specified SDN forwarding device let its attached sampling device (egHuawei's UTraffic or Netflow device) perform sampling, and also let the SDN forwarding device organize its controllable flow boot flow table entry (need to press the network) The results of the traffic engineering control module are done) and the uncontrolled flow (uncontrollable flow) of the original boot flow table entries.
- FIG. 4 is a schematic block diagram of a controller device in a communication network in accordance with one embodiment of the present invention. It should be understood that the controller 400 shown in FIG. 4 can implement various processes of the method of traffic engineering in the communication network involved in the embodiments of FIG. 2 and FIG. 3, and the operations and/or functions of the respective modules in the controller 400, respectively. For the corresponding processes in the method embodiments in FIG. 2 and FIG. 3, refer to the description in the foregoing method embodiments. To avoid repetition, the detailed description is omitted here.
- the communication network includes the controller and a plurality of nodes, the plurality of nodes including a flexible node and a non-flexible node, the flexible node being directly connected to the controller, and the non-flexible node is indirectly connected to the controller.
- the controller 400 shown in FIG. 4 includes a control unit 410, a transceiver unit 420, a determination unit 430, and a control unit 440.
- the control unit 410 is configured to determine a time frequency of the at least one sampling node and the traffic sample from the flexible node, where the transceiver unit 420 is configured to notify the at least one sampling node to perform traffic sampling at the time frequency, and receive all sampling nodes of each sampling.
- the traffic sampling data; the determining unit 430 is configured to use the traffic sampling data for each sampling and the total traffic of the link in the communication network that is higher than the first preset value or lower than the second preset value.
- the information determines parameter information required for traffic engineering; the control unit 440 is configured to perform traffic engineering control based on the parameter information.
- the embodiment of the present invention performs traffic sampling by using the determined at least one sampling node at a time frequency, and the controller notifies the at least one sampling node to sample data at a time frequency, and receives traffic sampling data of all sampling nodes of each sampling; According to the traffic, the data is used to determine the parameter information required by the traffic engineering, and finally the traffic engineering control is performed according to the parameter information. TE control under limited sampling is achieved.
- the traffic sampling data includes traffic flow in the communication network that passes through or starts from each sampling node to each destination node in the communication network.
- control unit is specifically configured to determine a time frequency of the at least one sampling node and the traffic sample according to the estimated total traffic of the service in the communication network.
- control unit is specifically configured to determine the number of the at least one sampling node and the time frequency according to the following formula:
- V*T s ⁇ 12 log(
- V*T s ⁇ 22
- represents the number of sampling nodes
- T s represents the sampling frequency
- V represents the estimated total traffic
- C represents a set of flexible nodes in the communication network
- represents a flexible node in the communication network.
- the number of ⁇ 11 , ⁇ 12 , ⁇ 21 , ⁇ 22 is a preset constant parameter, and 0 ⁇ 11 ⁇ 1, 0 ⁇ log( ⁇ 21
- the controller After determining the number of the sampling nodes, the controller randomly selects
- control unit is specifically configured to determine, for each destination node d, the number of the at least one sampling node and the time frequency according to the following formula:
- V d *T s d ⁇ 12 d log(
- the control unit for each of the destination node d After determining the number of samples that node, the control unit for each of the destination node d, the pre-generated based on the topology flexible nodes are selected in a directed acyclic graph DAG path
- nodes as the A set of sampling nodes, or randomly selecting
- the parameter information required by the traffic engineering includes injecting service traffic of each flexible node to each destination node in the communication network, and link utilization in the communication network is higher than a pre-pre Uncontrollable traffic in a set-valued link,
- the determining unit is specifically configured to determine parameter information required for traffic engineering according to the following formula (1):
- E represents all sets of links in the communication network
- N represents a set of all nodes in the communication network
- C represents a set of all flexible nodes in the communication network
- vec ( ⁇ I) means that all elements ⁇ I (for all u) are composed of vectors.
- e represents a link in the communication network whose link utilization is higher than the first preset value or a link lower than the second preset value
- f(e) represents the total traffic of the link e
- g(e ) indicates the amount of uncontrollable traffic in link e
- W wd represents the traffic flow in the communication network that passes through or starts from the sampling node w to the node d in the communication network
- I ud represents the injected service traffic from the flexible node u to the destination node d
- a packet is sent from node u to destination d, the total share of link e is denoted as ⁇ e (u,d), the total share of sampled node w is expressed as ⁇ w (u,d), and when w is u
- the outgoing package is expressed as u ⁇ w.
- the parameter information required for the traffic engineering includes injecting service traffic from each flexible node to each destination node in the communication network and link utilization in the communication network is higher than pre- Uncontrollable traffic in a set-valued link,
- the determining unit is specifically configured to determine parameter information required for traffic engineering according to any one of the following formulas (2), (3), and (4):
- E represents all sets of links in the communication network
- N represents a set of all nodes in the communication network
- C represents a set of all flexible nodes in the communication network
- vec ( ⁇ I) means that all elements ⁇ I (for all u) are composed of vectors.
- e represents a link in the communication network whose link utilization is higher than the first preset value or a link lower than the second preset value
- f(e) represents the total traffic of the link e
- g(e ) indicates the amount of uncontrollable traffic in link e
- W wd represents the traffic flow in the communication network that passes through or starts from the sampling node w to the node d in the communication network
- I ud represents the injected service traffic from the flexible node u to the destination node d
- a packet is sent from node u to destination d, the total share of link e is denoted as ⁇ e (u,d), the total share of sampled node w is expressed as ⁇ w (u,d), and when w is u
- the outgoing package is expressed as u ⁇ w;
- control unit is specifically configured to minimize the cost of all paths in the communication network according to the parameter information.
- control unit minimizes the cost of all paths in the communication network according to formula (5) or (6).
- control unit may be according to a specific example of (5), such as formula (5-11), formula (5-12), formula (5-13), and formula (5-14). And formula (5-15); and formula (5-21), formula (5-22), formula (5-23), formula (5-24), and formula (5-25); and formula (6)
- formulas (6-11) and formulas (6-12); and any of the formulas (6-21) and formulas (6-22) minimize the path cost.
- the present invention can implement fast TE control, and can implement existing TE control problems that are considered to be high complexity by different prior art technologies with low complexity algorithms.
- the embodiment of the present invention proposes and solves the more general TE control problem, so that the real-time TE can be implemented.
- FIG. 5 is a schematic block diagram of a controller device in a communication network according to another embodiment of the present invention.
- the controller 500 shown in FIG. 5 can implement various processes of the method of traffic engineering in the communication network involved in the embodiments of FIG. 2 and FIG. 3, and the operations and/or functions of the respective modules in the controller 500, respectively.
- the corresponding processes in the method embodiments in FIG. 2 and FIG. 3 refer to the description in the foregoing method embodiments. To avoid repetition, the detailed description is omitted here.
- the communication network includes the controller and a plurality of nodes, the plurality of nodes including a flexible node and a non-flexible node, the flexible node being directly connected to the controller, and the non-flexible node is indirectly connected to the controller.
- the controller 500 shown in FIG. 5 includes a processor 510 and a memory 520.
- the processor 500 can also include a bus system 530.
- the processor 510 and the memory 520 are connected by a bus system 530.
- the memory 520 is configured to store instructions
- the processor 510 is configured to execute an instruction stored by the memory 520 to determine a time frequency of the at least one sampling node and the traffic sample from the flexible node, and notify the at least one sampling node to perform traffic at the time frequency.
- the total traffic information of the link determines the parameter information required by the traffic engineering; the traffic engineering control is performed according to the parameter information.
- the embodiment of the present invention performs traffic sampling by using the determined at least one sampling node at a time frequency, and the controller notifies the at least one sampling node to sample data at a time frequency, and receives traffic sampling data of all sampling nodes of each sampling; According to the traffic, the data is used to determine the parameter information required by the traffic engineering, and finally the traffic engineering control is performed according to the parameter information. TE control under limited sampling is achieved.
- Processor 510 may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method may be completed by an integrated logic circuit of hardware in the processor 510 or an instruction in a form of software.
- the processor 510 may be a general-purpose processor, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), or the like. Programmable logic devices, discrete gates or transistor logic devices, discrete hardware components.
- DSP digital signal processor
- ASIC application specific integrated circuit
- FPGA field programmable gate array
- the general purpose processor may be a microprocessor or the processor or any conventional processor or the like.
- the steps of the method disclosed in the embodiments of the present invention may be directly implemented by the hardware decoding processor, or may be performed by a combination of hardware and software modules in the decoding processor.
- the software module can be located in a random access memory (RAM), a flash memory, a read-only memory (ROM), a programmable read only memory or an electrically erasable programmable memory, a register, etc. In the storage medium.
- the storage medium is located in the memory 520.
- the processor 510 reads the information in the memory 520 and completes the steps of the foregoing method in combination with hardware.
- the bus system 530 may include a power bus, a control bus, and a status signal bus in addition to the data bus. Wait. However, for clarity of description, various buses are labeled as bus system 530 in the figure.
- the traffic sampling data includes traffic flow in the communication network that passes through or starts from each sampling node to each destination node in the communication network.
- the processor 510 is specifically configured to determine a time frequency of the at least one sampling node and the traffic sample according to the estimated total traffic of the service in the communication network.
- the processor 510 is specifically configured to determine the number of the at least one sampling node and the time frequency according to the following formula:
- V*T s ⁇ 12 log(
- V*T s ⁇ 22
- represents the number of sampling nodes
- T s represents the sampling frequency
- V represents the estimated total traffic
- C represents a set of flexible nodes in the communication network
- represents a flexible node in the communication network.
- the number of ⁇ 11 , ⁇ 12 , ⁇ 21 , ⁇ 22 is a preset constant parameter, and 0 ⁇ 11 ⁇ 1, 0 ⁇ log( ⁇ 21
- the processor 510 randomly selects
- the processor 510 is specifically configured to determine, for each destination node d, the number of the at least one sampling node and the time frequency according to the following formula:
- V d *T s d ⁇ 12 d log(
- the processor 510 selects
- the parameter information required by the traffic engineering includes injecting service traffic of each flexible node to each destination node in the communication network, and link utilization in the communication network is higher than a pre-pre Uncontrollable traffic in a set-valued link,
- the processor 510 is specifically configured to determine parameter information required for traffic engineering according to formula (1).
- the parameter information required for the traffic engineering includes injecting service traffic from each flexible node to each destination node in the communication network and link utilization in the communication network is higher than pre- Uncontrollable traffic in a set-valued link,
- the processor 510 is specifically configured to determine parameter information required for traffic engineering according to any one of formula (2), formula (3), and formula (4).
- the processor 510 is specifically configured to minimize the cost of all paths in the communication network according to the parameter information.
- the processor 510 is specifically configured to minimize the cost of all paths in the communication network according to formula (5) or (6).
- control unit may be according to a specific example of (5), such as formula (5-11), formula (5-12), formula (5-13), and formula (5-14). And formula (5-15); and formula (5-21), formula (5-22), formula (5-23), formula (5-24), and formula (5-25); and formula (6)
- formulas (6-11) and formulas (6-12) and any of the formulas (6-21) and formulas (6-22) minimize the path cost. Specific For the sake of avoiding repetition, details are not described herein again.
- the present invention can implement fast TE control, and can implement existing TE control problems that are considered to be high complexity by different prior art technologies with low complexity algorithms.
- the embodiment of the present invention proposes and solves the more general TE control problem, so that the real-time TE can be implemented.
- system and “network” are used interchangeably herein.
- the term “and/or” in this context is merely an association describing the associated object, indicating that there may be three relationships, for example, A and / or B, which may indicate that A exists separately, and both A and B exist, respectively. B these three situations.
- the character "/" in this article generally indicates that the contextual object is an "or" relationship.
- B corresponding to A means that B is associated with A, and B can be determined according to A.
- determining B from A does not mean that B is only determined based on A, and that B can also be determined based on A and/or other information.
- the disclosed systems, devices, and methods may be implemented in other manners.
- the device embodiments described above are merely illustrative.
- the division of cells is only a logical function division.
- multiple units or components may be combined or integrated. Go to another system, or some features can be ignored or not executed.
- the mutual coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection through some interface, device or unit, or an electrical, mechanical or other form of connection.
- the units described as separate components may or may not be physically separate, and the components displayed as units may or may not be physical units, that is, may be located in one place, or may be distributed to multiple network units. Some or all of the units may be selected according to actual needs to achieve the objectives of the embodiments of the present invention.
- each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.
- the above integrated unit can be implemented in the form of hardware or in the form of a software functional unit.
- Computer readable media includes both computer storage media and communication media including any medium that facilitates transfer of a computer program from one location to another.
- a storage medium may be any available media that can be accessed by a computer.
- computer readable media may comprise RAM, ROM, EEPROM, CD-ROM or other optical disk storage, disk storage media or other magnetic storage device, or can be used for carrying or storing in the form of an instruction or data structure.
- the desired program code and any other medium that can be accessed by the computer may suitably be a computer readable medium.
- the software is transmitted from a website, server, or other remote source using coaxial cable, fiber optic cable, twisted pair, digital subscriber line (DSL), or wireless technologies such as infrared, radio, and microwave, then the coaxial cable , fiber optic cable, twisted pair, DSL, or wireless technologies such as infrared, wireless, and microwave are included in the fixing of the associated media.
- coaxial cable, fiber optic cable, twisted pair, DSL, or wireless technologies such as infrared, wireless, and microwave are included in the fixing of the associated media.
- a disk and a disc include a compact disc (CD), a laser disc, a compact disc, a digital versatile disc (DVD), a floppy disc, and a Blu-ray disc, wherein the disc is usually magnetically copied,
- the disc uses a laser to optically replicate the data. Combinations of the above should also be included within the scope of the computer readable media.
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Environmental & Geological Engineering (AREA)
- Software Systems (AREA)
- Algebra (AREA)
- Evolutionary Computation (AREA)
- Medical Informatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Artificial Intelligence (AREA)
- Physics & Mathematics (AREA)
- Databases & Information Systems (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Mathematical Physics (AREA)
- Probability & Statistics with Applications (AREA)
- Pure & Applied Mathematics (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
本发明实施例提供了一种用于通信网络中的流量工程的方法和控制器,该通信网络包括控制器和多个节点,该多个节点包括灵活节点和非灵活节点,该灵活节点与该控制器直接连接,该非灵活节点与该控制器间接相连,该方法包括:该控制器从该灵活节点中确定至少一个采样节点和流量采样的时间频率,该控制器通知该至少一个采样节点以该时间频率进行流量采样,并接收每一次采样的所有采样节点的流量采样数据;该控制器根据每一次采样的该流量采样数据和该通信网络中的链路利用率高于第一预设值或低于第二预设值的链路的总流量信息确定流量工程所需要的参数信息;该控制器根据该参数信息进行流量工程控制。本发明实施例能够实现有限采样下的TE控制。
Description
本申请要求于2016年02月05日提交中国专利局、申请号为201610081473.8、发明名称为“用于通信网络中的流量工程的方法和控制器”的中国专利申请的优先权,其全部内容通过引用结合在本申请中。
本申请涉及通信领域,特别涉及一种用于通信网络中的流量工程的方法和控制器。
网络流量工程(Traffic Engineering,TE)是一种方法来控制网络流量,以获得网络整体性的某些特定目标,如优化网络利用率,均衡化网络利用率,减少网络拥塞程度,减少网络负载的技术。
要达到网络的这种整体性目标,TE的最常用的实现方法是先对全网中所有源和目的地连接需求进行估计(或者统计),进一步利用各种路由机制(或网络协议),根据TE目标进行优化。随着网络规模日益增长,原有的TE的方法计算复杂度巨大,要执行TE的所有步骤,其可行性很低。举例来说,在互联网中使用的最典型的路由协议都是基于以下两种机理产生的:
距离向量(Distance Vector)路由协议,例如,路由信息协议(Routing Information Protocol,RIP),边界网关协议(Border Gateway ProtocolBGP);和链路状态(Link State)路由协议,例如,开放式最短路径优先协议(Open Shortest Path First,OSPF)。
然而,现有的路由协议的计算复杂度可以是很大的(特别是对于大规模网络),如距离向量路由使用贝尔曼福特(Bellman-Ford)算法,有O(|V|*|E|)的复杂性,其中|V|是节点数(number of node),以及|E|是边数(number of edge);而链路状态协议通常使用一些Dijkstra算法的变种,有O(|E|+|V|log|V|)的复杂性,在这些路由协议中,计算最好的路径的方法是把短期链路成本设置为离目的地的距离有关的成本,但TE还需要进行多次反复运行这些算法去更新这短期链路成本,去实现前述的某些特定的全网的控制目标,可见其复杂度是很大的。实际应用时,网络流量工程师一般会手动试错法去更新这短期链路成本,该流程是非常复杂的,在大规模网络上是不可能实施的。现有的一些TE是用多协议标签交换(Multiprotocol Label Switching,MPLS)方法去实现,然而,需要建立链路标签的成本,尽管MPLS反复运行的不再是现有的路由协议,但要实现TE优化目标,亦需要反复更新标签。尽管这仅是一些例子,实际的TE实现取决于网络流量工程师,例如在选择网络流量工程师可以引入上述协议的某些变体,如改变链路成本或目标,或允许多个路径,以实现其TE的目标,当中要反复运行这些算法亦是不能避免的。随着网络的规模的升级,如果要完成TE的每一步,即由估计全网连路流量需求,到按TE目标做全网路由优化,原有TE的巨大复杂性已不再适用。
再者,网络流量工程所要的全网连路流量需求,是基于网络状态数据而估计的,如果运行TE的算法速度慢,即使网络状态数据采样频率高,随着网络状态的改变,TE所用的有效网络状态信息实际上是有限的(稀疏的)。然而,现有TE技术都是基于网络状态数据是充足而且是不变的假设。这种只有稀疏的网络状态数据的情况,按照原有的基
于全网络的算法难以实现TE控制。
基本上,所有网络都有TE问题,由于软件定义网络(Software Defined Network,SDN)是一个未来网络的重要的可建结构,下面以SDN网络举例说明现有的TE问题。SDN把控制平面(control plane)跟数据平面(Data plane)分离。因SDN控制器对网络的可控性较强(因跟以前分布式网络不同,SDN采用集中式控制器的结构(architecture)),所以SDN上的TE的效能都比现有网络为佳。但要达到SDN TE的(潜在)最佳效能(至少要优于现有网络的TE效能),现有技术是同样是需要全网状态的。
就SDN网络而言。SDN的结构是集中式的(使得其TE的效能都比现有网络为佳),在进行TE控制前,需要估计全网流量需求。但如果要将全网络TE所需的相关讯息都采样,并反馈(feedback)到SDN控制器上,所需的测量数据反馈开销会很大。而且这些信息的有效性一般都低于TE的运算时间(因为正如前文所述,SDN控制器的高运算复杂度因为所需运算量大,在有运算结果前,网络场景起变化是正常的)。并且,现有的SDN的技术需要完全的网络讯息(故一般都要密集式采样),还有需要假设这些网络讯息在TE Control的运算过程中保持为不变量(即常数)。按照原有的基于全网络的算法难以实现TE控制。
为了减少大量不必要的采样(measurement)和反馈(feedback),并能实时实现TE目标,如何在低采样和低反馈开销的环境下,同时建立一些一般性的而又可用到这些低采样和低反馈的数据并可以通过低复杂度算法实现的TE优化题,做到等同于全网SDN在全可观环境下TE的预期效果成为重中之重的问题。
因此,在现有网络下,如何实现在有限采样下的TE成为亟待解决的问题。
发明内容
本发明实施例提供一种用于通信网络中的流量工程的方法和控制器,该方法能够实现有限采样下的TE。
第一方面,提供了一种用于通信网络中的流量工程的方法,该通信网络包括控制器和多个节点,该多个节点包括灵活节点和非灵活节点,该灵活节点与该控制器直接连接,该非灵活节点与该控制器间接相连,该方法包括:
该控制器从该灵活节点中确定至少一个采样节点和流量采样的时间频率,
该控制器通知该至少一个采样节点以该时间频率进行流量采样,并接收每一次采样的所有采样节点的流量采样数据;
该控制器根据每一次采样的该流量采样数据和该通信网络中的链路利用率高于第一预设值或低于第二预设值的链路的总流量信息确定流量工程所需要的参数信息;
该控制器根据该参数信息进行流量工程控制。
因此,通过确定出的至少一个采样节点以时间频率进行流量采样,该控制器通知该至少一个采样节点以时间频率采样数据,并接收每一次采样的所有采样节点的流量采样数据;并根据该流量采用数据确定流量工程所需要的参数信息,最后根据该参数信息进行流量工程控制。实现了在有限采样下的高效高性能TE控制。
应理解,该流量采样数据包括该通信网络中经过或起始于每一个采样节点到该通信网络中的每一个目的节点的业务流量。
应理解,控制器可以按照现有的通知采用节点,例如可以通过流表下发的方式等,
本发明实施例并不对此做限定,同样的,也可以按照现有的方式实现控制器获取采用节点发送的采样数据,本发明实施例并不对此做限定。
流量工程控制的目的是实现一些TE目标。例如:降低网络拥塞程度、优化网络利用率、平衡网络负载等。控制器可以通过降低通信网络中的链路的最大利用率来实现部分的TE目标。
结合第一方面,在第一方面的一种实现方式中,该流量采样数据包括该通信网络中经过或起始于每一个采样节点到该通信网络中的每一个目的节点的业务流量。
结合第一方面及其上述实现方式,在第一方面的另一种实现方式中,该控制器从该灵活节点中确定至少一个采样节点和流量采样的时间频率,包括:
该控制器根据该通信网络中业务的预估总流量确定该至少一个采样节点和流量采样的时间频率。
结合第一方面及其上述实现方式,在第一方面的另一种实现方式中,该控制器根据该通信网络中业务的预估总流量确定该至少一个采样节点和流量采样的时间频率,包括:
该控制器根据以下公式确定该至少一个采样节点的个数和该时间频率:
|Cs|=γ11|C|
V*Ts=γ12log(|Cs|)
或者
|Cs|=log(γ21|C|)
V*Ts=γ22|Cs|,
其中,|Cs|表示该采样节点的个数,Ts表示该采样频率,V表示该预估总流量,C表示该通信网络中灵活节点的集合,|C|表示该通信网络中灵活节点的个数,γ11,γ12,γ21,γ22为预设常数参数,且0<γ11<1,0<log(γ21|C|)<|C|,
在确定该采样节点个数后,该控制器在该灵活节点集合中随机选取|Cs|个节点作为该采样节点的集合;
应理解,在本发明实施例中,随机的机率不受限制性,例如可以是同等机率,也可以是按节点的被访问的历史次数的比例作为机率的决定,也可以是按节点的度的比例作为机率的决定,本发明实施例并不对此做限定。
或者,该控制器,对每一个目的节点d,根据以下公式确定该至少一个采样节点的个数和该时间频率:
|Cs
d|=γ11
d|C|
Vd*Ts
d=γ12
dlog(|Cs
d|)
或者
|Cs
d|=log(γ21
d|C|)
Vd*Ts
d=γ22
d|Cs
d|,
其中,|Cs
d|表示对应该目的节点d的采样节点的个数,Ts
d表示对应该目的节点d的采样频率,Vd表示到该目的节点d的预估总流量,C表示该通信网络中灵活节点的集合,|C|表示该d通信网络中灵活节点的个数,γ11
d,γ12
d,γ21
d,γ22
d为预设常数参数,且0<γ11
d<1,0<log(γ21
d|C|)<|C|,
在确定该采样节点个数后,该控制器针对每一个目的节点d,在预先基于灵活节点的拓扑生成的有向非循环图DAG中的路径中选取的|Cs
d|个节点,作为该采样节点的集合,
或在所述灵活节点集合中随机选取|Cs
d|个节点,作为所述采样节点的集合。
例如每一个选取的采样节点都是从DAG中的独立的路径中选取出来的,例如从所有目的节点d的DAG共同的部分中选取共同的|Cs|个节点,换句话说,每一个目的节点d的DAG图可以为同一个DAG图,也可以为不同的DAG图,本发明实施例并不限于此。
结合第一方面及其上述实现方式,在第一方面的另一种实现方式中,该流量工程所需要的参数信息包括每一个灵活节点到该通信网络中的每一个目的节点的注入业务流量和该通信网络中的链路利用率高于预设值的链路中不可控的业务量,
该控制器根据每一次采样的该流量采样数据和该通信网络中的链路利用率高于第一预设值或低于第二预设值的链路的总流量信息确定流量工程所需要的参数信息,包括根据以下公式(1)确定流量工程所需要的参数信息:
限制条件为:
Icurrent=Iud=ΔI+Ilast iteration
其中,E表示该通信网络中所有链路集合,N表示该通信网络中所有节点的集合,C表示该通信网络中所有灵活节点的集合;
其中,w表示采样节点,u表示灵活节点,d表示目的节点,ΔI是指上次采样时灵活节点u的注入流量Ilast iteration与当前采样时灵活节点u的注入流量Iud的差,vec(ΔI)表示把所有元件ΔI(对于所有的u)组成向量。
e表示该通信网络中的链路利用率高于第一预设值的一个链路或低于第二预设值的一个链路,f(e)表示链路e的总流量,g(e)表示链路e中不可控的业务量,
Wwd表示该通信网络中经过或起始于采样节点w到该通信网络中的节点d的业务流量,Iud表示灵活节点u到目的节点d的注入业务流量,
从节点u发出一个包去目的地d,经过链路e的总份额表示为αe(u,d),经过采样节点w的总份额表示为βw(u,d),且当w被u发出的包经过表示成u≤w。
上述描述了通过求解l0范数最小化确定流量工程所需要的参数信息,本发明实施例中,还可以通过其他替代的方式来确定流程工程所需要的参数信息,例如,可以通过求解l1范数最小化、加权l1范数最小化或者通过求解lp范数最小化来确定流量工程所需要的参数信息,其中,0<p≤1。下面分别进行详细说明。
结合第一方面及其上述实现方式,在第一方面的另一种实现方式中,该流量工程所需要的参数信息包括每一个灵活节点到该通信网络中的每一个目的节点的注入业务流量和该通信网络中的链路利用率高于预设值的链路中不可控的业务量,
该控制器根据每一次采样的该流量采样数据和该通信网络中的链路利用率高于第一预设值或低于第二预设值的链路的总流量信息确定流量工程所需要的参数信息,包括根据以下公式(2)、公式(3)和公式(4)中的任意一个确定流量工程所需要的参数信息:
限制条件为:
Icurrent=Iud=ΔI+Ilast iteration
其中,E表示该通信网络中所有链路集合,N表示该通信网络中所有灵活节点的集合,C表示该通信网络中所有灵活节点的集合;
其中,w表示采样节点,u表示灵活节点,d表示目的节点,ΔI是指上次采样时灵活节点u的注入流量Ilast iteration与当前采样时灵活节点u的注入流量Iud的差,vec(ΔI)表示把所有元件ΔI(对于所有的u)组成向量。
e表示该通信网络中的链路利用率高于第一预设值的一个链路或低于第二预设值的一个链路,f(e)表示链路e的总流量,g(e)表示链路e中不可控的业务量,
Wwd表示该通信网络中经过或起始于采样节点w到该通信网络中的节点d的业务流量,Iud表示灵活节点u到目的节点d的注入业务流量,
从节点u发出一个包去目的地d,经过链路e的总份额表示为αe(u,d),经过采样节点w的总份额表示为βw(u,d),且当w被u发出的包经过表示成u≤w;
限制条件为:
Icurrent=Iud=ΔI+Ilast iteration
限制条件为:
Icurrent=Iud=ΔI+Ilast iteration
结合第一方面及其上述实现方式,在第一方面的另一种实现方式中,所述控制器根
据每一次采样的所述流量采样数据和所述通信网络中的链路利用率高于第一预设值或低于第二预设值的链路的总流量信息确定流量工程所需要的参数信息,包括根据凸优化方法(其中还可以包括剃度算法)求解公式(2)和公式(3),和以连续凸优化逼近法(Continuous convex optimization approximation)(其中可以包括连续使用剃度算法)求解公式(4)中的任意一个确定流量工程所需要的参数信息。
结合第一方面及其上述实现方式,在第一方面的另一种实现方式中,该控制器根据该参数信息进行流量工程控制,包括:该控制器根据该参数信息最小化所述通信网络中的所有路径的成本。
结合第一方面及其上述实现方式,在第一方面的另一种实现方式中,该控制器根据该参数信息最小化所述通信网络中的所有路径的成本,包括该控制器根据以下公式(5)所述通信网络的网络成本:
限制条件为:
其中,e表示所述链路,E表示所有链路的集合,C表示灵活节点的集合,u表示灵活节点,d表示目的节点,N表示所有节点的集合,Iud表示灵活节点u到目的节点d的注入业务流量,Pud表示所有从u至d的可行路径的集合,
应注意,可行路径的示例是按传统的方法,例如在OSPF(开放式最短路径优先协议)中,如果u’是路径中的一点而且是传统节点(非灵活节点),路径P中u’的下一节点便是往d的最短路径中的下一节点,可行路径的另一个示例可以是网络管理员自定义的路径集合。
P∈Pud表示P表示Pud中的一条可行路径。x(P)表示路径P的流量。表示所有从u至d的可行路径的总流量。ρ(e)表示链路e的可控带宽的利用率,即其中c(e)表示链路e的总带宽,g(e)表示链路e的不可控流的流量,表示链路e上的所有可行路径的可控流的总流量;表示网络成本;
即按所有链路e的可控带宽的利用率构造的一个函数,即设定所有链路e的可控带宽的利用率会给出一个代表成本的实数,例如成本便是所有链路的最大利用率;又例如成本便是把路径P的成本以γP做加权后的总和,当中可以是即路径P上链路e的成本以γe做加权后的总和,当中可以是链路利用率ρ(e),公式(5)便表示最小化所述通信网络中的(加权后)最拥堵的地方;当中也可以是链路利用率的倒数,即1/ρ(e),公式(5)便表示最小化所述通信网络中的(加权后)最不均衡或最不被利用的地方。
结合第一方面及其上述实现方式,在第一方面的另一种实现方式中,所述控制器根据所述参数信息最小化所述通信网络中的所有路径的成本,包括所述控制器根据以下公式(6)最小化所述通信网络中的所有路径的最大成本:
限制条件为:
其中,e表示所述链路,E表示所有链路的集合,C表示灵活节点的集合,u表示灵活节点,d表示目的节点,N表示所有节点的集合,Iud表示灵活节点u到目的节点d的注入业务流量,Pud表示所有从u至d的可行路径的集合,P∈Pud表示P表示Pud中的一条可行路径。x(P)表示路径P的流量。表示所有从u至d的可行路径的总流量,ρ(e)表示链路e的可控带宽的利用率,即其中c(e)表示链路e的总带宽,g(e)表示链路e的不可控流的流量,表示链路e上的所有可行路径的可控流的总流量;表示路径P的成本。
结合第一方面及其上述实现方式,在第一方面的另一种实现方式中,所述控制器根据所述参数信息最小化所述通信网络中所有路径的成本,包括所述控制器根据公式(5-11)最小化所述通信网络中的网络成本:
限制条件为:
0≤he(θ)≤1
其中,e表示所述链路,E表示所有链路的集合,C表示灵活节点的集合,u表示灵活节点,d表示目的节点,N表示所有节点的集合,Iud表示灵活节点u到目的节点d的注入业务流量,Pud表示所有从u至d的可行路径的集合,P∈Pud表示P表示Pud中的一条可行路径,x(P)表示路径P的流量,λ表示最大的网络吞吐量比例,表示所有从u至d的可行路径的总流量,θ表示链路的最大利用率相关参数,he(θ)用来表示权重参数与θ的关系函数,其中权重参数包括γe和γP,γe表示链路e的权重参数,γP路径P的权重参数,c(e)表示链路e的总带宽,g(e)表示链路e的不可控流的流量,表示链路e上的所有可行路径的可控流的总流量,
其中,根据权重参数γe和γP的不同权重情况,he(θ)可以为以下中的任意一种:
也就是说,在构造表示网络成本的时,要考虑权重参数设定不同权重情况,对应的he(θ)具有不同的形式。如中有一单调递增子函数对所有链路的ρ(e)都是一样时,这子函数可被更换为另一单调递增子函数,路径P的权重参数γP>0可以乘积方法或以指数方法加权在链路e的权重参数γe>0也可以乘积方法或以指数方法加权在所以这里有四种权重双参数(γP,γe)设定方法:两者都是乘积法时,两者都是指数法时,或者(γP用乘积法,γe用指数法)时,或者(γP用指数法,γe用乘积法)时,或者例如
本发明实施例可以取例如:本发明实施例可以取例如:
本发明实施例可以取例如:本发明实施例可
以取不失一般性的,当中可以是因为可被更换为另一单调递增子函数,如
针对上述四种情况,公式(5-11)可以变换成如下公式(5-12)、(5-13)、(5-14)和(5-15)。
限制条件为:
限制条件为:
限制条件为:
限制条件为:
结合第一方面及其上述实现方式,在第一方面的另一种实现方式中,所述控制器根
据所述参数信息最小化所述通信网络中所有路径的成本,包括所述控制器根据公式(5-21)最小化所述通信网络中的网络成本:
限制条件为:
0≤he(θ)≤1
其中,e表示所述链路,E表示所有链路的集合,C表示灵活节点的集合,u表示灵活节点,d表示目的节点,N表示所有节点的集合,Iud表示灵活节点u到目的节点d的注入业务流量,Pud表示所有从u至d的可行路径的集合,P∈Pud表示P表示Pud中的一条可行路径,x(P)表示路径P的流量,λ表示最大的网络吞吐量比例,表示所有从u至d的可行路径的总流量,θ表示链路的均衡相关参数,he(θ)用来表示权重参数与θ的关系函数,其中权重参数包括γe和γP,γe表示链路e的权重参数,γP路径P的权重参数,c(e)表示链路e的总带宽,g(e)表示链路e的不可控流的流量,表示链路e上的所有可行路径的可控流的总流量,
其中,根据权重参数γe和γP的不同权重情况,he(θ)可以为以下中的任意一种:
也就是说,在构造表示网络成本时,要考虑权重参数设定不同权重情况,对应的he(θ)具有不同的形式。如中有一单调递增子函数对所有链路的ρ(e)都是一样时,这子函数可被更换为另一单调递增子函数,路径P的权重参数γP>0可以乘积方法或以指数方法加权在链路e的权重参数γe>0也可以乘积方法或
以指数方法加权在所以这里有四种权重双参数(γP,γe)设定方法:两者都是乘积法时,两者都是指数法时,或者(γP用乘积法,γe用指数法)时,或者(γP用指数法,γe用乘积法)时,或者例如
本发明实施例可以取例如:本发明实施例可以取例如:
本发明实施例可以取例如:本发明实施例可以取不失一般性的,当中可以是因为可被更换为另一单调递增子函数,如
针对上述四种情况,公式(5-21)可以变换成如下公式(5-22)、(5-23)(5-24)和(5-25)。
限制条件为:
限制条件为:
限制条件为:
限制条件为:
1≤θ
结合第一方面及其上述实现方式,在第一方面的另一种实现方式中,所述控制器根据所述参数信息最小化所述通信网络中所有路径的成本,包括所述控制器根据公式(6-11)最小化所述通信网络中的网络成本:
限制条件为:
0≤ke(λ)≤λ
其中,e表示所述链路,E表示所有链路的集合,C表示灵活节点的集合,u表示灵活节点,d表示目的节点,N表示所有节点的集合,Iud表示灵活节点u到目的节点d的注入业务流量,Pud表示所有从u至d的可行路径的集合,P∈Pud表示P表示Pud中的一条可行路径,x(P)表示路径P的流量,表示所有从u至d的可行路径的总流量,λ表示最大的网络吞吐量相关参数,ke(λ)用于表示链路e的权重参数与λ的对应关系,ρ(e)表示链路e的可控带宽的利用率,c(e)表示链路e的总带宽,g(e)表示链路e的不可控流的流量,表示链路e上的所有可行路径的可控流的总流量。
结合第一方面及其上述实现方式,在第一方面的另一种实现方式中,所述控制器根式(6-11)最小化所述通信网络中的网络成本,包括:
所述控制器将所述ke(λ)作为常数,并采用低复杂度的最大并发分数流算法求解公式(6-11)最小化所述通信网络中的网络成本。
也就是说,控制器可以根据把(6-11)中的ke(λ)看成一个常数(即当前迭代参数),以低复杂度的最大并发分数流算法最小化所述通信网络中的网络成本。其中,最大并发分数流算法中的短期链路成本会跟这当前迭代参数有关,例如,短期路径成本与链路的利用率成正比或正相关。
结合第一方面及其上述实现方式,在第一方面的另一种实现方式中,
所述控制器根据所述参数信息最小化所述通信网络中所有路径的成本,包括所述控制器根据公式(6-12)最小化所述通信网络中的所有路径的最大成本:
限制条件为:
0≤ke(λ)≤λ
其中,e表示所述链路,E表示所有链路的集合,C表示灵活节点的集合,u表示灵活节点,d表示目的节点,N表示所有节点的集合,Iud表示灵活节点u到目的节点d的注入业务流量,Pud表示所有从u至d的可行路径的集合,P∈Pud表示P表示Pud中的一条可行路径,x(P)表示路径P的流量,表示所有从u至d的可行路径的总流量,λ表示最大的网络吞吐量相关参数,ke(λ)用于表示链路e的权重参数与λ的对应关系,其中γe表示链路的权重参数,ρ(e)表示链路e的可控带宽的利用率,c(e)表示链路e的总带宽,g(e)表示链路e的不可控流的流量,表示链路e上的所有可行路径的可控流的总流量,其中,根据权重参数γe不同权重情况,ke(λ)可以为以下中的任意一种:
结合第一方面及其上述实现方式,在第一方面的另一种实现方式中,
所述控制器以下公式:公式(6-12)最小化所述通信网络中的网络成本,包括:
所述控制将ke(λ)作为常数,并采用低复杂度的最大并发分数流算法求解公式(6-12)最小化所述通信网络中的所有路径的最大成本。
也就是说,控制器可以根据把(6-12)中的ke(λ)看成一个常数(即当前迭代参数),以低复杂度的最大并发分数流算法最小化所述通信网络中的网络成本。其中,最大并发分数流算法中的短期链路成本会跟这当前迭代参数有关,例如,短期路径成本与链路的利用率成正比或正相关。
结合第一方面及其上述实现方式,在第一方面的另一种实现方式中,所述控制器根据所述参数信息最小化所述通信网络中所有路径的成本,包括所述控制器根据公式(6-21)最小化所述通信网络中的网络成本:
限制条件为:
0≤ke(λ)≤1
其中,e表示所述链路,E表示所有链路的集合,C表示灵活节点的集合,u表示灵活节点,d表示目的节点,N表示所有节点的集合,Iud表示灵活节点u到目的节点d的注入业务流量,Pud表示所有从u至d的可行路径的集合,P∈Pud表示P表示Pud中的一条可行路径,x(P)表示路径P的流量,λ表示均衡网络下的网络吞吐量相关参数,
表示所有从u至d的可行路径的总流量,ke(λ)用于表示链路e的权重参数与λ的对应关系,其中权重参数包括γe和γP,γe表示链路权重参数,γP表示路径权重参数),ρ(e)表示链路e的可控带宽的利用率,c(e)表示链路e的总带宽,g(e)表示链路e的不可控流的流量,表示链路e上的所有可行路径的可控流的总流量。
结合第一方面及其上述实现方式,在第一方面的另一种实现方式中,所述控制器公式(6-21)最小化所述通信网络中的网络成本,包括:
所述控制器将所述ke(λ)作为常数,并采用低复杂度的最大并发分数流算法(Fractional Maximum Concurrent Flow)求解公式(6-21)最小化所述通信网络中的网络成本。
也就是说,控制器可以根据把(6-21)中的ke(λ)看成一个常数(即当前迭代参数),即采取其任意一种,以低复杂度的最大并发分数流算法最小化所述通信网络中的网络成本。其中,最大并发分数流算法中的短期链路成本会跟这当前迭代参数有关,例如,短期路径成本与链路的利用率成正比或正相关。
结合第一方面及其上述实现方式,在第一方面的另一种实现方式中,所述控制器根据所述参数信息最小化所述通信网络中所有路径的成本,包括所述控制器根据公式(6-22)最小化所述通信网络中的所有路径的最大成本:
限制条件为:
0≤ke(λ)≤1
其中,e表示所述链路,E表示所有链路的集合,C表示灵活节点的集合,u表示灵活节点,d表示目的节点,N表示所有节点的集合,Iud表示灵活节点u到目的节点d的注入业务流量,Pud表示所有从u至d的可行路径的集合,P∈Pud表示P表示Pud中的一条可行路径,x(P)表示路径P的流量,λ表示均衡网络下的网络吞吐量相关参数,表示所有从u至d的可行路径的总流量,ke(λ)用于表示链路e的权重参数与λ的对应关系,其中权重参数包括γe和γP,γe表示链路权重参数,γP表示路径权重参数,ρ(e)表示链路e的可控带宽的利用率,c(e)表示链路e的总带宽,g(e)表示链路e的不可控流的流量,表示链路e上的所有可行路径的可控流的总流量;其中,根据权重参数γe不同权重情况,ke(λ)可以为以下中的任意一种:
ke(λ)=γeλ;
结合第一方面及其上述实现方式,在第一方面的另一种实现方式中,所述控制器根据公式(6-22)最小化所述通信网络中的网络成本,包括:
所述控制器将所述ke(λ)作为常数,并采用改进的最大并发分数流算法求解公式(6-22)最小化所述通信网络中的网络成本,
也就是说,控制器根据把(6-22)中的ke(λ)看成一个常数(即当前迭代参数),以一个改版的最大并发分数流算法(即把原版的最大并发分数流算法中的短期链路成本看成正值,在此之上,设置一个新的跟这当前迭代参数有关的短期链路成本,并把它看成负值)最小化所述通信网络中的网络成本。
因此,通过上述方案,本发明实例,能够实现快速TE控制,并且能够以低复杂度的算法去实现不同现有技术认为高复杂度的现有TE控制问题。换句话说,本发明实施例提出并解决更普遍性的TE控制问题,使实时TE可以实现。
第二方面,提供了一种通信网络中的控制器,该控制器能够实现第一方面及其实现方式中的任一实现方式,该控制器中的各个模块的操作和/或功能,分别用于实现的第一方面及其实现方式中的相应方法特征,为了简洁,在此不再赘述。
第三方面,提供了一种通信网络中的控制器,该控制器包括存储指令的存储器和处理器,其中,该处理器执行该指令进行如第一方面及其各种实现方式中的任一种用于通信网络中的流量工程的方法。
第四方面,提供了一种处理装置,该处理装置应用于通信系统中。该处理装置可以为一个或多个处理器或芯片。在其他可能情况下,该处理装置也可以为通信系统中的实体装置或虚拟装置。该处理装置被配置用于执行上述第一方面及其各种实现方式中的任一种用于通信网络中的流量工程的方法。
第五方面,提供了一种计算机程序产品,该计算机程序产品包括:计算机程序代码,当该计算机程序代码被通信设备的计算单元、处理单元或处理器运行时,使得该通信设备执行上述第一方面及其各种实现方式中的任一种用于通信网络中的流量工程的方法。
第六方面,提供了一种计算机可读存储介质,该计算机可读存储介质存储有程序,该程序使得通信设备执行上述第一方面及其各种实现方式中的任一种用于通信网络中的流量工程的方法。
基于上述技术方案,本发明实施例通过确定出的至少一个采样节点以时间频率进行流量采样,该控制器通知该至少一个采样节点以时间频率采样数据,并接收每一次采样的所有采样节点的流量采样数据;并根据该流量采用数据确定流量工程所需要的参数信息,最后根据该参数信息进行流量工程控制。实现了在有限采样下的TE控制。
图1是本发明实施例可使用的SDN网络系统示意图。
图2是根据本发明一个实施例的用于通信网络中的流量工程的方法示意流程图。
图3是根据本发明另一实施例的用于通信网络中的流量工程的方法示意流程图。
图4是根据本发明一个实施例的通信网络中控制器设备的示意框图。
图5是根据本发明另一实施例的通信网络中控制器设备的示意框图。
下面将结合本发明实施例中的附图,对本发明实施例中的技术方案进行清楚、完整地描述。
应理解,本发明实例可以应用于各种通信网络(尤指通信系统的网络层),例如:IP网络(IP network),基于多协议标签交换网络(MPLS-based network),自组织网络(Ad-hoc network),信息中心网络(Information Centric Network),内容分发网络(Content Distribution Network),SDN,而且应理解,本发明实施例的技术方案可以兼容于各种通信系统的技术(尤指通信系统的接入层)(即不依赖于特定通信系统的技术)例如可以兼容以下通信系统的技术:不同的无线通信系统,包括全球移动通讯(Global System of Mobile communication,GSM)系统、码分多址(Code Division Multiple Access,CDMA)系统、宽带码分多址(Wideband Code Division Multiple Access,WCDMA)系统、通用分组无线业务(General Packet Radio Service,GPRS)、长期演进(Long Term Evolution,LTE)系统、LTE频分双工(Frequency Division Duplex,FDD)系统、LTE时分双工(Time Division Duplex,TDD)、通用移动通信系统(Universal Mobile Telecommunication System,UMTS)或全球互联微波接入(Worldwide Interoperability for Microwave Access,WiMAX)通信系统及固定线路网络(简称固网)等。下面仅以SDN网络(及以固网技术),举例进行详细说明。
图1是本发明实施例可使用的SDN网络系统示意图。如图1所示的SDN网络架构中包括控制器和节点1至15,其中,节点2、9、14直接与控制器连接,为灵活节点;节点1、3、4、5、6、7、8、10、11、12、13、15间接与该控制器间接相连,为非灵活节点。
下面结合图1的场景,首先定义对本发明实施例中的一些跟TE有关的流量数据的术语和符号:
在图1中,该SDN网络可以被建模为一个图G=(N,E),所有节点的集合被表示为N,而所有链路的集合被表示为E;在集合N中,当中有一子集合C,称为SDN转发设备(SDN-Forwarding Element,简称SDN-FE;即灵活节点(Flexible Node)或1类节点)),它们是直接连接到SDN控制器(SDN Controller,简称SDN-)的节点;其他节点均称为非灵活节点或2类节点,所组成的集合被表示为D.即灵活节点集合为C,和非灵活节集合为D。在图1中,节点上的号码代表节点的编号,集合N包含所有节点,即节点1到节点15,而集合C包括{2,9,14},这三个节点,集合D包括{1,3,4,5,6,7,8,10,11,12,13,15}。
从源节点行进到目的地节点而不经过灵活节点的流被称为不可控流(Uncontrollable Traffic),
其他业务必须经过至少一个灵活节点(或始发于灵活节点(即灵活节点是业务的源节点)),因灵活节点的可控性,这些业务也称可控流(Controllable Traffic)。
对应于每一链路e,总流量表示为f(e),不可控的部分表示为g(e)。
Tsd表示从N中的源节点s到N中的目的节点的总流量。
Wud表示经过或起始与C中的节点u到N中的目的节点d的总流量。
Iud表示由C中的节点u注入到到N中的目的节点d的总流量。
根据上述定义,可以肯定的是Wud≥Tud.。
图2为根据本发明一个实施例的用于通信网络中的流量工程的方法,该通信网络例如可以为如图1中的SDN网络,本发明实施例并不限于此,该通信网络包括控制器和多个节点,该多个节点包括灵活节点和非灵活节点,该灵活节点与该控制器直接连接,该非灵活节点与该控制器间接相连,该方法可以由控制器执行,该方法包括:
210,控制器从该灵活节点中确定至少一个采样节点和流量采样的时间频率。
220,该控制器通知该至少一个采样节点以该时间频率进行流量采样,并接收每一次采样的所有采样节点的流量采样数据。
230,该控制器根据每一次采样的该流量采样数据和该通信网络中的链路利用率高于第一预设值或低于第二预设值的链路的总流量信息确定流量工程所需要的参数信息。
240,该控制器根据该参数信息进行流量工程控制。
因此,本发明实施例通过确定出的至少一个采样节点以时间频率进行流量采样,该控制器通知该至少一个采样节点以时间频率采样数据,并接收每一次采样的所有采样节点的流量采样数据;并根据该流量采用数据确定流量工程所需要的参数信息,最后根据该参数信息进行流量工程控制。实现了在有限采样下的TE控制。
应理解,应理解第一预设预设大于第二预设值。该流量采样数据包括该通信网络中经过或起始于每一个采样节点到该通信网络中的每一个目的节点的业务流量。
应理解,在220中,控制器可以按照现有的通知采用节点,例如可以通过流表下发的方式等,本发明实施例并不对此做限定,同样的,也可以按照现有的方式实现控制器获取采用节点发送的采样数据,本发明实施例并不对此做限定。
可选地,作为另一实施例,在210中,该控制器根据该通信网络中业务的预估总流量确定该至少一个采样节点和流量采样的时间频率。
进一步地,作为另一实施例,在210中,该控制器根据以下公式确定该至少一个采样节点的个数和该时间频率:
|Cs|=γ11|C|
V*Ts=γ12log(|Cs|)
或者
|Cs|=log(γ21|C|)
V*Ts=γ22|Cs|,
其中,|Cs|表示该采样节点的个数,Ts表示该采样频率,V表示该预估总流量,C表示该通信网络中灵活节点的集合,|C|表示该通信网络中灵活节点的个数,γ11,γ12,γ21,γ22为预设常数参数,且,0<γ11<1,0<log(γ21|C|)<|C|,
在确定该采样节点个数后,该控制器在该灵活节点集合中随机选取|Cs|个节点作为该采样节点的集合。
应理解,在本发明实施例中,随机的机率不受限制性,例如可以是同等机率,也可以是按节点的被访问的历史次数的比例作为机率的决定,也可以是按节点的度的比例作为机率的决定,本发明实施例并不对此做限定。
或者,该控制器,对每一个目的节点d,根据以下公式确定该至少一个采样节点的个数和该时间频率:
|Cs
d|=γ11
d|C|
Vd*Ts
d=γ12
dlog(|Cs
d|)
或者
|Cs
d|=log(γ21
d|C|)
Vd*Ts
d=γ22
d|Cs
d|,
其中,|Cs
d|表示对应该目的节点d的采样节点的个数,Ts
d表示对应该目的节点d的采样频率,Vd表示到该目的节点d的预估总流量,C表示该通信网络中灵活节点的集合,|C|表示该d通信网络中灵活节点的个数,γ11
d,γ12
d,γ21
d,γ22
d为预设常数参数,且0<γ11<1,0<log(γ21
d|C|)<|C|。
该控制器针对每一个目的节点d,在预先基于灵活节点的拓扑生成的有向非循环图(Directed Acyclic Graph,DAG)中的路径中选取的|Cs
d|个节点,作为该采样节点的集合,或在所述灵活节点集合中随机选取|Cs
d|个节点,作为所述采样节点的集合。
例如每一个选取的采样节点都是从DAG中的独立的路径中选取出来的,例如从所有目的节点d的DAG共同的部分中选取共同的|Cs
d|个节点,换句话说,每一个目的节点d的DAG图可以为同一个DAG图,也可以为不同的DAG图,本发明实施例并不限于此。
可选地,作为另一实施例,该流量工程所需要的参数信息包括每一个灵活节点到该通信网络中的每一个目的节点的注入业务流量和该通信网络中的链路利用率高于预设值的链路中不可控的业务量,
在230中,控制器根据以下公式(1)确定流量工程所需要的参数信息:
限制条件为:
Icurrent=Iud=ΔI+Ilast iteration
其中,E表示该通信网络中所有链路集合,N表示该通信网络中所有节点的集合,C表示该通信网络中所有灵活节点的集合;
其中,w表示采样节点,u表示灵活节点,d表示目的节点,ΔI是指上次采样时灵活节点u的注入流量Ilast iteration与当前采样时灵活节点u的注入流量Iud的差,vec(ΔI)表示把所有元件ΔI(对于所有的u)组成向量。表示vec(ΔI)的最小l0范数,e表示该通信网络中的链路利用率高于第一预设值的一个链路或低于第二预设值的一个链路,f(e)表示链路e的总流量,g(e)表示链路e中不可控的业务量,Wwd表示该通信网络中经过或起始于采样节点w到该通信网络中的节点d的业务流量,Iud表示灵活节点u到目的节点d的注入业务流量,
从节点u发出一个包去目的地d,经过链路e的总份额表示为αe(u,d),经过采样节点w的总份额表示为βw(u,d),且当w被u发出的包经过表示成u≤w。
上述描述了通过求解l0范数最小化确定流量工程所需要的参数信息,本发明实施例中,还可以通过其他替代的方式来确定流程工程所需要的参数信息,例如,可以通过求解l1范数最小化、加权l1范数最小化或者通过求解lp范数最小化来确定流量工程所需要
的参数信息,其中,0<p≤1。下面分别进行详细说明。
可替代的,在230中,控制器根据以下公式(2)、公式(3)和公式(4)中的任意一个确定流量工程所需要的参数信息:
其中,公式(2)表示求解l1范数最小化,即
限制条件为:
Icurrent=Iud=ΔI+Ilast iteration
其中,E表示该通信网络中所有链路集合,N表示该通信网络中所有灵活节点的集合,C表示该通信网络中所有灵活节点的集合;
其中,w表示采样节点,u表示灵活节点,d表示目的节点,ΔI是指上次采样时灵活节点u的注入流量Ilast iteration与当前采样时灵活节点u的注入流量Iud的差,vec(ΔI)表示把所有元件ΔI(对于所有的u)组成向量。
e表示该通信网络中的链路利用率高于第一预设值的一个链路或低于第二预设值的一个链路,f(e)表示链路e的总流量,g(e)表示链路e中不可控的业务量,
Wwd表示该通信网络中经过或起始于采样节点w到该通信网络中的节点d的业务流量,Iud表示灵活节点u到目的节点d的注入业务流量,
从节点u发出一个包去目的地d,经过链路e的总份额表示为αe(u,d),经过采样节点w的总份额表示为βw(u,d),且当w被u发出的包经过表示成u≤w;MinΔI||vec(ΔI)||l1表示vec(ΔI)的最小l1范数。
公式(3)表示求解加权l1范数最小化,即
初始化设定权重m=1
重复求解以下公式(3),直到一些(可以选择)优化参数不变或几乎没有变化,例如w,或ΔI不变或几乎没有变化,然后停止:
限制条件为:
Icurrent=Iud=ΔI+Ilast iteration
公式(4)表示求解lp范数最小化,或加权lp范数最小化(其中0<p≤1,即把公式(2),或公式(3)的l1范数更改成lp范数.
限制条件为:
Icurrent=Iud=ΔI+Ilast iteration
应理解,在240中,流量工程控制的目的是实现一些TE目标。例如:降低网络拥塞程度、优化网络利用率、平衡网络负载等。控制器可以通过降低通信网络中的链路的最大利用率来实现部分的TE目标。
相应地,作为另一实施例,在240中,该控制器根据该参数信息最小化所述通信网络中的所有路径的成本。
进一步地,作为另一实施例,在240中,该控制器根据以下公式(5)最小化所述通信网络的网络成本:
限制条件为:
其中,e表示所述链路,E表示所有链路的集合,C表示灵活节点的集合,u表示灵活节点,d表示目的节点,N表示所有节点的集合,Iud表示灵活节点u到目的节点d的注入业务流量,Pud表示所有从u至d的可行路径的集合,应注意,可行路径的示例是按传统的方法,例如在开放式最短路径优先协议(OSPF)中,如果u’是路径中的一点而且是传统节点(非灵活节点),路径P中u’的下一节点便是往d的最短路径中的下一节点,可行路径的另一个示例可以是网络管理员自定义的路径集合。
P∈Pud表示P表示Pud中的一条可行路径。x(P)表示路径P的流量。表示所有从u至d的可行路径的总流量,ρ(e)表示链路e的可控带宽的利用率,即其中c(e)表示链路e的总带宽,g(e)表示链路e的不可控流的流量,表示链路e上的所有可行路径的可控流的总流量;表示网络成本。即按路径P上所有链路e的可控带宽的利用率构造的一个函数,即设定路径P上所有链路e的可控带宽的利用率会给出一个代表成本的实数,例如成本便是路径中链路的最大利用率。
可替代的,作为另一实施例,控制器根据以下公式(6)最小化所述通信网络中的所有路径的最大成本:
限制条件为:
其中,e表示所述链路,E表示所有链路的集合,C表示灵活节点的集合,u表示灵活节点,d表示目的节点,N表示所有节点的集合,Iud表示灵活节点u到目的节点d的注入业务流量,Pud表示所有从u至d的可行路径的集合,P∈Pud表示P表示Pud中的一条可行路径。x(P)表示路径P的流量。表示所有从u至d的可行路径的总流量,ρ(e)表示链路e的可控带宽的利用率,即其中c(e)表示链路e的总带宽,g(e)表示链路e的不可控流的流量,表示链路e上的所有可行路径的可控流的总流量;表示路径P的成本。
应理解,可替代地,该控制器根据该参数信息,根据公式(5)或(6)的等量转换的解或近似解及其实现方法,具体地,可以根据(5)的特殊例子,如公式(5-11)、公式(5-12)、公式(5-13)、公式(5-14)和公式(5-15);和公式(5-21)、公式(5-22)、公式(5-23)、公式(5-24)和公式(5-25);以及公式(6)的特殊例子公式(6-11)和公式(6-12);以及公式(6-21)和公式(6-22)中任一种公式最小化路径成本。下面分别进行描述。
可替代地,控制器根据以下公式:公式(5-11)、公式(5-12)、公式(5-13)、公式(5-14)和公式(5-15)中的任意一种最小化所述通信网络中的网络成本。
具体地,所述控制器根据所述参数信息最小化所述通信网络中所有路径的成本,包括所述控制器根据公式(5-11)最小化所述通信网络中的网络成本:
限制条件为:
0≤he(θ)≤1
其中,e表示所述链路,E表示所有链路的集合,C表示灵活节点的集合,u表示灵活节点,d表示目的节点,N表示所有节点的集合,Iud表示灵活节点u到目的节点d的注入业务流量,Pud表示所有从u至d的可行路径的集合,P∈Pud表示P表示Pud中的一条可行路径,x(P)表示路径P的流量,表示所有从u至d的可行路径的总流量,λ表示最大的网络吞吐量比例,θ表示链路的最大利用率相关参数,he(θ)用来表示权重参数与θ的关系函数,其中权重参数包括γe和γP,γe表示链路e的权重参数,γP路径P
的权重参数,c(e)表示链路e的总带宽,g(e)表示链路e的不可控流的流量,表示链路e上的所有可行路径的可控流的总流量,
其中,根据权重参数γe和γP的不同权重情况,he(θ)可以为以下中的任意一种:
也就是说,在构造表示网络成本的时,要考虑权重参数设定不同权重情况,对应的he(θ)具有不同的形式。如中有一单调递增子函数对所有链路的ρ(e)都是一样时,这子函数可被更换为另一单调递增子函数,路径P的权重参数γP>0可以乘积方法或以指数方法加权在链路e的权重参数γe>0也可以乘积方法或以指数方法加权在所以这里有四种权重双参数(γP,γe)设定方法:两者都是乘积法时,两者都是指数法时,或者(γP用乘积法,γe用指数法)时,或者(γP用指数法,γe用乘积法)时,或者例如
本发明实施例可以取例如:本发明实施例可以取例如:
本发明实施例可以取例如:本发明实施例可以取不失一般性的,当中可以是因为可被更换为另一单调递增子函数,如
针对上述四种情况,公式(5-11)可以变换成如下公式(5-12)、(5-13)、(5-14)和(5-15)。
限制条件为:
限制条件为:
限制条件为:
限制条件为:
可替代地,控制器根据以下公式:公式(5-21)、公式(5-22)、公式(5-23)、公式(5-24)和公式(5-25)中的任意一种最小化所述通信网络中的网络成本:
具体地,所述控制器根据所述参数信息最小化所述通信网络中所有路径的成本,包括所述控制器根据公式(5-21)最小化所述通信网络中的网络成本:
限制条件为:
0≤he(θ)≤1
其中,e表示所述链路,E表示所有链路的集合,C表示灵活节点的集合,u表示灵活节点,d表示目的节点,N表示所有节点的集合,Iud表示灵活节点u到目的节点d的注入业务流量,Pud表示所有从u至d的可行路径的集合,P∈Pud表示P表示Pud中的一条可行路径,x(P)表示路径P的流量,λ表示最大的网络吞吐量比例,表示所有从u至d的可行路径的总流量,θ表示链路的均衡相关参数,he(θ)用来表示权重参数与θ的关系函数,其中权重参数包括γe和γP,γe表示链路e的权重参数,γP路径P的权重参数,c(e)表示链路e的总带宽,g(e)表示链路e的不可控流的流量,表示链路e上的所有可行路径的可控流的总流量,
其中,根据权重参数γe和γP的不同权重情况,he(θ)可以为以下中的任意一种:
也就是说,在构造表示网络成本时,要考虑权重参数设定不同权重情况,对应的he(θ)具有不同的形式。如中有一单调递增子函数对所有链路的ρ(e)都是一样时,这子函数可被更换为另一单调递增子函数,路径P的权重参数γP>0可以乘积方法或以指数方法加权在链路e的权重参数γe>0也可以乘积方法或以指数方法加权在所以这里有四种权重双参数(γP,γe)设定方法:两者都是乘积法时,两者都是指数法时,或者(γP用乘积法,γe用指数法)时,或者(γP用指数法,γe用乘积法)时,或者例如
本发明实施例可以取例如:本发明实施例可以取例如:
本发明实施例可以取例如:本发明实施例可以取不失一般性的,当中可以是因为可被更换为另一单调递增子函数,如
针对上述四种情况,公式(5-21)可以变换成如下公式(5-22)、(5-23)(5-24)和(5-25)。
限制条件为:
限制条件为:
限制条件为:
限制条件为:
1≤θ
可替代地,控制器根据以下公式:公式(6-11)和公式(6-12)中的任意一种最小化所述通信网络中的所有路径的网络成本。
具体地,所述控制器根据所述参数信息最小化所述通信网络中所有路径的成本,包括所述控制器根据公式(6-11)最小化所述通信网络中的网络成本:
限制条件为:
0≤ke(λ)≤λ
其中,e表示所述链路,E表示所有链路的集合,C表示灵活节点的集合,u表示灵活节点,d表示目的节点,N表示所有节点的集合,Iud表示灵活节点u到目的节点d的注入业务流量,Pud表示所有从u至d的可行路径的集合,P∈Pud表示P表示Pud中的一条可行路径,x(P)表示路径P的流量,表示所有从u至d的可行路径的总流量,λ表示最大的网络吞吐量相关参数,ke(λ)用于表示链路e的权重参数与λ的对应关系,ρ(e)表示链路e的可控带宽的利用率,c(e)表示链路e的总带宽,g(e)表示链路e的不可控流的流量,表示链路e上的所有可行路径的可控流的总流量。
进一步地,作为另一实施例,所述控制器根式(6-11)最小化所述通信网络中的网络成本,包括:
所述控制器将所述ke(λ)作为常数,并采用低复杂度的最大并发分数流算法求解公式(6-11)最小化所述通信网络中的网络成本。
也就是说,控制器可以根据把(6-11)中的ke(λ)看成一个常数(即当前迭代参数),以低复杂度的最大并发分数流算法最小化所述通信网络中的网络成本。其中,最大并发分数流算法中的短期链路成本会跟这当前迭代参数有关,例如,短期路径成本与链路的利用率成正比或正相关。
可替代地,所述控制器根据所述参数信息最小化所述通信网络中所有路径的成本,包括所述控制器根据公式(6-12)最小化所述通信网络中的所有路径的最大成本:
限制条件为:
0≤ke(λ)≤λ
其中,e表示所述链路,E表示所有链路的集合,C表示灵活节点的集合,u表示灵活节点,d表示目的节点,N表示所有节点的集合,Iud表示灵活节点u到目的节点d的注入业务流量,Pud表示所有从u至d的可行路径的集合,P∈Pud表示P表示Pud中的一条可行路径,x(P)表示路径P的流量,表示所有从u至d的可行路径的总流量,λ表示最大的网络吞吐量相关参数,ke(λ)用于表示链路e的权重参数与λ的对应关系,其中γe表示链路的权重参数,ρ(e)表示链路e的可控带宽的利用率,c(e)表示链路e的总带宽,g(e)表示链路e的不可控流的流量,表示链路e上的所有可行路径的可控流的总流量,其中,根据权重参数γe不同权重情况,ke(λ)可以为以下中的任意一种:
进一步地,作为另一实施例,所述控制器以下公式:公式(6-12)最小化所述通信网络中的网络成本,包括:
所述控制将ke(λ)作为常数,并采用低复杂度的最大并发分数流算法求解公式(6-12)最小化所述通信网络中的所有路径的最大成本。
也就是说,控制器可以根据把(6-12)是的ke(λ)看成一个常数(即当前迭代参数),以低复杂度的最大并发分数流算法最小化所述通信网络中的网络成本。其中,最大并发分数流算法中的短期链路成本会跟这当前迭代参数有关,例如,短期路径成本与链路的利用率成正比或正相关。
可替代地,控制器根据以下公式:公式(6-21)和公式(6-22)中的任意一种最小化所述通信网络中的所有路径的最大成本。
具体地,所述控制器根据所述参数信息最小化所述通信网络中所有路径的成本,包括所述控制器根据公式(6-21)最小化所述通信网络中的网络成本:
限制条件为:
0≤ke(λ)≤1
其中,e表示所述链路,E表示所有链路的集合,C表示灵活节点的集合,u表示灵活节点,d表示目的节点,N表示所有节点的集合,Iud表示灵活节点u到目的节点d的注入业务流量,Pud表示所有从u至d的可行路径的集合,P∈Pud表示P表示Pud中的一条可行路径,x(P)表示路径P的流量,λ表示均衡网络下的网络吞吐量相关参数,表示所有从u至d的可行路径的总流量,ke(λ)用于表示链路e的权重参数与λ的对应关系,其中权重参数包括γe和γP,γe表示链路权重参数,γP表示路径权重参数,ρ(e)表示链路e的可控带宽的利用率,c(e)表示链路e的总带宽,g(e)表示链路e的不可控流的流量,表示链路e上的所有可行路径的可控流的总流量。
进一步地,作为另一实施例所述控制器根据公式(6-21)最小化所述通信网络中的网络成本,包括:
所述控制器将所述ke(λ)作为常数,并采用改进的最大并发分数流算法求解公式(6-21)最小化所述通信网络中的网络成本。
也就是说,控制器可以根据把(6-21)中的ke(λ)看成一个常数(即当前迭代参数),即采取其任意一种,以低复杂度的最大并发分数流算法最小化所述通信网络中的网络成本。其中,最大并发分数流算法中的短期链路成本会跟这当前迭代参数有关,例如,短期路径成本与链路的利用率成正比或正相关。
可替代地,所述控制器根据所述参数信息最小化所述通信网络中所有路径的成本,包括所述控制器根据公式(6-22)最小化所述通信网络中的所有路径的最大成本:
限制条件为:
0≤ke(λ)≤1
其中,e表示所述链路,E表示所有链路的集合,C表示灵活节点的集合,u表示灵活节点,d表示目的节点,N表示所有节点的集合,Iud表示灵活节点u到目的节点d的
注入业务流量,Pud表示所有从u至d的可行路径的集合,P∈Pud表示P表示Pud中的一条可行路径,x(P)表示路径P的流量,λ表示均衡网络下的网络吞吐量相关参数,表示所有从u至d的可行路径的总流量,ke(λ)用于表示链路e的权重参数与λ的对应关系,其中权重参数包括γe和γP,γe表示链路权重参数,γP表示路径权重参数,ρ(e)表示链路e的可控带宽的利用率,c(e)表示链路e的总带宽,g(e)表示链路e的不可控流的流量,表示链路e上的所有可行路径的可控流的总流量;其中,根据权重参数γe不同权重情况,ke(λ)可以为以下中的任意一种:
ke(λ)=γeλ;
进一步地,作为另一实施例所述控制器根据公式(6-22)最小化所述通信网络中的网络成本,包括:
所述控制将ke(λ)作为常数,并采用改进的最大并发分数流算法求解公式(6-22)最小化所述通信网络中的所有路径的最大成本。
也就是说,控制器根据把(6-22)中的ke(λ)看成一个常数(即当前迭代参数),以一个改版的最大并发分数流算法(即把原版的最大并发分数流算法中的短期链路成本看成正值,在此之上,设置一个新的跟这当前迭代参数有关的短期链路成本,并把它看成负值)最小化所述通信网络中的网络成本。
因此,通过上述方案,本发明实例,能够实现快速TE控制,并且能够以低复杂度的算法去实现不同现有技术认为高复杂度的现有TE控制问题。换句话说,本发明实施例提出并解决更普遍性的TE控制问题,使实时TE可以实现。
应理解,本领域技术人员根据所给出的图2的例子,显然可以进行各种等价的修改或变化,这样的修改或变化也落入本发明实施例的范围内。例如,在实际应用中,可以按照图2的方法进行实时或非实时的流量工作控制,本发明实施例并不对此做限定。再例如,图2中仅以SDN网络为例描述了本发明实施例的流量工程的方法,但本发明实施例并不限于SDN网络中。也就是说,本发明实施例的流量工程的方法还可以应用在其他网络中,当应用于其他网络时,上述方法仍然适用,例如,应用于其他网络中,由于没有灵活节点和非灵活节点之分,即统称为节点,因此,只要将本发明实施例中的灵活节点看成网络中的节点即可实现流程工程控制,这样的修改也在本发明的保护范围内。
下面结合图3的具体地例子,进一步详细描述本发明实施例的流量工程的方法。应注意,图3的例子仅仅是为了帮助本领域技术人员理解本发明实施例,而非要将本发明实施例限于所例示的具体数值或具体场景。本领域技术人员根据所给出的图3的例子,显然可以进行各种等价的修改或变化,这样的修改或变化也落入本发明实施例的范围内。
图3所示的流量工程主要分为三个大的过程:第一过程:采样/测量(Measurement);这里为有限度采样(Partial Measurement)。
第二过程:网络流量模型(Network Traffic Model)(for TE input recovery);网络流量模
型主要功能为恢复输入参数;即利用第一步的结果去恢复第三步控制过程(control process)所需的输入参数。因为第一步的结果通常都不是控制过程所需要的输入参数。
第三过程:控制过程(Control Process)。其主要功能是通过第二步中所恢复出的网络流量有关的输入参数,来进行流量工程,即透过全网路由优化,以达到TE目标。
其中,图3中的310-340对应上述第一过程,350对应上述第二过程,360对应上述第三过程。
下面对各个过程分别进行详细描述,具体地,图3所示的方法可以由控制器中个各个功能模块执行,应注意,本发明实施例中为了描述的方便,每一个步骤都可以由对应的功能模块来执行。这里的各个功能模块只是为了描述方便而划分的,在实际应用中,由相应的实体单元来实现相应的功能模块的功能,例如,各个功能模块都可以是处理器,可以由处理器产生指令完成实现功能模块的功能。本发明实施例并不对此限定。下面对各个步骤分别进行详细描述。具体地该方法300包括:
310,网络拓扑识别。
该过程可以由拓扑识别模块(Topology identification Module)执行,本发明的描述都是基于拓扑是已知的假设。其中,一种获得拓扑的方法是透过链路层发现协议(Link layer Discovery Protocol(LLDP))。本发明实施例并不对此做限定。
320,网络流量预测。
该过程可以由采样频率和定位模块(Sampling Frequency and Location Module)执行,本发明实施例中,可选先做一次流量测量,取得在一个时间单位内的流量总和(先从拓扑识别模块知道网络节点位置和链路),或做一次初步估计,例如用历史数据,作为初步估计。下文会详加展述。
330,确定采样频率和采样节点。
该过程可以由采样频率和定位模块(Sampling Frequency and Location Module)执行,采样频率和定位模块会根据310获得网络节点位置和链路,进而根据网络流量预测模块所预测的流量,去决定采样频率(换言之,决定采样时距是多少个时间单位)和(在一个采样时距中的)采样位置。
在指定的时刻,SDN控制器将该流量采样需求(其决定的采样比率),告诉SDN转发设备(SDN-FE)的相关取样装置。
具体而言,该控制器(例如可以是采样频率和定位模块)根据以下公式确定该至少一个采样节点的个数和该时间频率:
|Cs|=γ11|C|
V*Ts=γ12log(|Cs|)
或者
|Cs|=log(γ21|C|)
V*Ts=γ22|Cs|,
其中,|Cs|表示该采样节点的个数,Ts表示该采样频率,V表示该预估总流量,C表示该通信网络中灵活节点的集合,|C|表示该通信网络中灵活节点的个数,γ11,γ12,γ21,γ22为预设常数参数,且,0<γ11<1,0<log(γ21|C|)<|C|,
在确定该采样节点个数后,该控制器在该灵活节点集合中随机选取|Cs|个节点作为该采样节点的集合,
应理解,在本发明实施例中,随机的机率不受限制性,例如可以是同等机率,也可以是按节点的被访问的历史次数的比例作为机率的决定,也可以是按节点的度的比例作为机率的决定,本发明实施例并不对此做限定。
或者,该控制器,对每一个目的节点d,根据以下公式确定该至少一个采样节点的个数和该时间频率:
|Cs
d|=γ11
d|C|
Vd*Ts
d=γ12
dlog(|Cs
d|)
或者
|Cs
d|=log(γ21
d|C|)
Vd*Ts
d=γ22
d|Cs
d|,
其中,|Cs
d|表示对应该目的节点d的采样节点的个数,Ts
d表示对应该目的节点d的采样频率,Vd表示到该目的节点d的预估总流量,C表示该通信网络中灵活节点的集合,|C|表示该d通信网络中灵活节点的个数,γ11
d,γ12
d,γ21
d,γ22
d为预设常数参数,且0<γ11
d<1,0<log(γ21
d|C|)<|C|.
该控制器针对每一个目的节点d,在预先基于灵活节点的拓扑生成的有向非循环图(Directed Acyclic Graph,DAG)中的路径中选取的|Cs
d|个节点,作为该采样节点的集合,或在所述灵活节点集合中随机选取|Cs
d|个节点,作为所述采样节点的集合。
例如每一个选取的采样节点都是从DAG中的独立的路径中选取出来的,例如从所有目的节点d的DAG共同的部分中选取共同的|Cs
d|个节点,换句话说,每一个目的节点d的DAG图可以为同一个DAG图,也可以为不同的DAG图,本发明实施例并不限于此。
340,获取测量数据。
该过程可以由采样(测量)模块(Sampling Module)执行,为了流量工程(TE)的目标,对一些跟TE有关的数据进行采样。在SDN问题上,本发明需要按采样频率和定位模块所指定的时间和采用节点的位置,在每一个采样时距的时间中,对经过流W进行采样。具体的采用过程,可以按照现有的方法进行,本发明并不对具体的采样方法做限定。
采样节点会根据采样频率采样位置去采样。在指定的时刻,在指定的采用节点,具体的采样方法和流程,可采用现有技术执行,例如,可以按照如下方法进行,但本发明实施例并不限于此。
控制器将流表条目信息和采样节点信息进行匹配,生成组表条目(当中包括采样比率,以执行采样行为)和引导流表条目(包括流信息),并将其发送给采样节点,以使采用节点按照采用频率采样。
最后,采样节点将采样的测量数据发送给控制器,即控制器获得测量数据。
350,确定流量工程所需要的参数信息。
该过程可以由输入恢复模块(Input Recovery Module)执行,输入恢复模块从测量数据,推断出需要(或可以)用于流量控制(TE)的因素的值.(它的每一次运作时间是以一个采样时距为单位。)
它的输入是可用的网络测量结果(即采样模块的输出结果);在实施例中,在一个时间单位中,它是对于每一个目的地d和每一个SDN-FE节点u的经过流,W(u,d)和每一条链路e的总流量total traffic,f(e).即在一个采样时距中,对每一个d,W和f(e)是它的输入。
它的输出是TE可用的网络状态;在实施例中,在一个时间单位中,它是对于每一个目的地d和每一个SDN-FE节点u的注入流,I(u,d)和每一条链路e的uncontrollable traffic,g(e)。即在一个采样时距中,对每一个d,I和g是它的输出。
具体而言,控制器可以根据公式(1)确定流量工程所需要的参数信息:
上述描述了通过求解l0范数最小化确定流量工程所需要的参数信息,本发明实施例中,还可以通过其他替代的方式来确定流程工程所需要的参数信息,例如,可以通过求解l1范数最小化、加权l1范数最小化或者通过求解lp范数最小化来确定流量工程所需要的参数信息,其中,0<p≤1。下面分别进行详细说明。
也就是说,控制器可以根据公式(2)、公式(3)和公式(4)中的任意一个确定流量工程所需要的参数信息。
360,流量工程控制。
该过程可以由网络流量工程控制模块(TE Control Module)执行,这个模块的目标是是实现一些TE目标。例如:降低网络拥塞程度、优化网络利用率、平衡网络负载等。控制器可以通过降低通信网络中的链路的最大利用率来实现部分的TE目标。
应理解,流程工程中的每一次运作时间是可以一个时间单位或一个采样时距为单位。它的输入是可用的网络状态(即输入恢复模块的输出结果)。它的输出是对于每一个灵活节点u和目的地节点d,它都给出从u到d每一条路径和其流量。
具体而言,该控制器根据该参数信息最小化所述通信网络中的所有路径的成本。
进一步地,该控制器根据公式(5)或(6)最小化所述通信网络中的所有路径的成本。
可替代的,作为另一实施例,控制器根据可以根据(5)的特殊例子,如公式(5-11)、公式(5-12)、公式(5-13)、公式(5-14)和公式(5-15);和公式(5-21)、公式(5-22)、公式(5-23)、公式(5-24)和公式(5-25);以及公式(6)的特殊例子公式(6-11)和公式(6-12;以及公式(6-21)和公式(6-22)中任一种公式最小化路径成本,具体地,可以参照对应方法实施例的各个过程,为避免重复,此处不再赘述。
应注意,本发明实施例中的控制器还可以包括其他模块,例如,通讯模块(Communication Module)主要负责各模块之间的通讯,所以可有也可没有,因其功能也可以放进其他模块。
在指定的时刻,这模块会令SDN控制器将采样频率和定位模块所指定的流量采样需求(包括采样比率),和网络流量工程控制模块所制造的流表信息(包括引导流表条目),发送到指定的SDN转发设备,让其附属的采样设备(e.g.Huawei’s UTraffic or Netflow device),执行采样,也可让SDN转发设备组织其可控流量(controllable flow)的引导流表条目(需要按网络流量工程控制模块的结果去做)和不可控流量(uncontrollable flow)的原版的引导流表条目.
应理解,上述各过程的序号的大小并不意味着执行顺序的先后,各过程的执行顺序应以其功能和内在逻辑确定,而不应对本发明实施例的实施过程构成任何限定。
图4是根据本发明一个实施例的通信网络中控制器设备的示意框图。应理解,图4所示的控制器400能够实现图2和图3实施例中涉及的通信网络中的流量工程的方法的各个过程,控制器400中的各个模块的操作和/或功能,分别为了实现图2和图3中的方法实施例中的相应流程,具体可参见上述方法实施例中的描述,为避免重复,此处适当省略详述描述。
具体的,该通信网络包括该控制器和多个节点,该多个节点包括灵活节点和非灵活节点,该灵活节点与该控制器直接连接,该非灵活节点与该控制器间接相连。图4所示的控制器400包括:控制单元410、收发单元420、确定单元430和控制单元440。
控制单元410用于从该灵活节点中确定至少一个采样节点和流量采样的时间频率,收发单元420用于通知该至少一个采样节点以该时间频率进行流量采样,并接收每一次采样的所有采样节点的流量采样数据;确定单元430用于根据每一次采样的该流量采样数据和该通信网络中的链路利用率高于第一预设值或低于第二预设值的链路的总流量信息确定流量工程所需要的参数信息;控制单元440用于根据该参数信息进行流量工程控制。
因此,本发明实施例通过确定出的至少一个采样节点以时间频率进行流量采样,该控制器通知该至少一个采样节点以时间频率采样数据,并接收每一次采样的所有采样节点的流量采样数据;并根据该流量采用数据确定流量工程所需要的参数信息,最后根据该参数信息进行流量工程控制。实现了在有限采样下的TE控制。
可选地,作为另一实施例,该流量采样数据包括该通信网络中经过或起始于每一个采样节点到该通信网络中的每一个目的节点的业务流量。
可选地,作为另一实施例,该控制单元具体用于根据该通信网络中业务的预估总流量确定该至少一个采样节点和流量采样的时间频率。
进一步地,作为另一实施例,该控制单元具体用于根据以下公式确定该至少一个采样节点的个数和该时间频率:
|Cs|=γ11|C|
V*Ts=γ12log(|Cs|)
或者
|Cs|=log(γ21|C|)
V*Ts=γ22|Cs|,
其中,|Cs|表示该采样节点的个数,Ts表示该采样频率,V表示该预估总流量,C表示该通信网络中灵活节点的集合,|C|表示该通信网络中灵活节点的个数,γ11,γ12,γ21,γ22为预设常数参数,且0<γ11<1,0<log(γ21|C|)<|C|,
在确定该采样节点个数后,该控制器在该灵活节点集合中随机选取|Cs|个节点作为该采样节点的集合;
或者,该控制单元具体用于对每一个目的节点d,根据以下公式确定该至少一个采样节点的个数和该时间频率:
|Cs
d|=γ11
d|C|
Vd*Ts
d=γ12
dlog(|Cs
d|)
或者
|Cs
d|=log(γ21
d|C|)
Vd*Ts
d=γ22
d|Cs
d|,
其中,|Cs
d|表示对应该目的节点d的采样节点的个数,Ts
d表示对应该目的节点d的采样频率,Vd表示到该目的节点d的预估总流量,C表示该通信网络中灵活节点的集合,|C|表示该d通信网络中灵活节点的个数,γ11
d,γ12
d,γ21
d,γ22
d为预设常数参数,且0<γ11
d<1,0<log(γ21
d|C|)<|C|,
在确定该采样节点个数后,该控制单元针对每一个目的节点d,在预先基于灵活节点的拓扑生成的有向非循环图DAG中的路径中选取的|Cs
d|个节点,作为该采样节点的集合,或在所述灵活节点集合中随机选取|Cs
d|个节点,作为所述采样节点的集合。
可选地,作为另一实施例,该流量工程所需要的参数信息包括每一个灵活节点到该通信网络中的每一个目的节点的注入业务流量和该通信网络中的链路利用率高于预设值的链路中不可控的业务量,
该确定单元具体用于根据以下公式(1)确定流量工程所需要的参数信息:
限制条件为:
Icurrent=Iud=ΔI+Ilast iteration
其中,E表示该通信网络中所有链路集合,N表示该通信网络中所有节点的集合,C表示该通信网络中所有灵活节点的集合;
其中,w表示采样节点,u表示灵活节点,d表示目的节点,ΔI是指上次采样时灵活节点u的注入流量Ilast iteration与当前采样时灵活节点u的注入流量Iud的差,vec(ΔI)表示把所有元件ΔI(对于所有的u)组成向量。
e表示该通信网络中的链路利用率高于第一预设值的一个链路或低于第二预设值的一个链路,f(e)表示链路e的总流量,g(e)表示链路e中不可控的业务量,
Wwd表示该通信网络中经过或起始于采样节点w到该通信网络中的节点d的业务流量,Iud表示灵活节点u到目的节点d的注入业务流量,
从节点u发出一个包去目的地d,经过链路e的总份额表示为αe(u,d),经过采样节点w的总份额表示为βw(u,d),且当w被u发出的包经过表示成u≤w。
可替代地,作为另一实施例,该流量工程所需要的参数信息包括每一个灵活节点到该通信网络中的每一个目的节点的注入业务流量和该通信网络中的链路利用率高于预设值的链路中不可控的业务量,
该确定单元具体用于根据以下公式(2)、公式(3)和公式(4)中的任意一种确定流量工程所需要的参数信息:
限制条件为:
Wcurrent=βIcurrent
Icurrent=Iud=ΔI+Ilast iteration
其中,E表示该通信网络中所有链路集合,N表示该通信网络中所有节点的集合,C表示该通信网络中所有灵活节点的集合;
其中,w表示采样节点,u表示灵活节点,d表示目的节点,ΔI是指上次采样时灵活节点u的注入流量Ilast iteration与当前采样时灵活节点u的注入流量Iud的差,vec(ΔI)表示
把所有元件ΔI(对于所有的u)组成向量。
e表示该通信网络中的链路利用率高于第一预设值的一个链路或低于第二预设值的一个链路,f(e)表示链路e的总流量,g(e)表示链路e中不可控的业务量,
Wwd表示该通信网络中经过或起始于采样节点w到该通信网络中的节点d的业务流量,Iud表示灵活节点u到目的节点d的注入业务流量,
从节点u发出一个包去目的地d,经过链路e的总份额表示为αe(u,d),经过采样节点w的总份额表示为βw(u,d),且当w被u发出的包经过表示成u≤w;
限制条件为:
Wcurrent=βIcurrent
Icurrent=Iud=ΔI+Ilast iteration
限制条件为:
Wcurrent=βIcurrent
Icurrent=Iud=ΔI+Ilast iteration
可选地,作为另一实施例,该控制单元具体用于根据该参数信息最小化所述通信网络中的所有路径的成本。
进一步地,作为另一实施例,该控制单元,根据公式(5)或(6)最小化所述通信网络中的所有路径的成本。
可替代的,作为另一实施例,该控制单元可以根据(5)的特殊例子,如公式(5-11)、公式(5-12)、公式(5-13)、公式(5-14)和公式(5-15);和公式(5-21)、公式(5-22)、公式(5-23)、公式(5-24)和公式(5-25);以及公式(6)的特殊例子公式(6-11)和公式(6-12);以及公式(6-21)和公式(6-22)中任一种公式最小化路径成本。具体地,可以参照对应方法实施例的各个过程,为避免重复,此处不再赘述。
因此,通过上述方案,本发明实例,能够实现快速TE控制,并且能够以低复杂度的算法去实现不同现有技术认为高复杂度的现有TE控制问题。换句话说,本发明实施例提出并解决更普遍性的TE控制问题,使实时TE可以实现。
图5是根据本发明另一实施例的通信网络中控制器设备的示意框图。应理解,图5所示的控制器500能够实现图2和图3实施例中涉及的通信网络中的流量工程的方法的各个过程,控制器500中的各个模块的操作和/或功能,分别为了实现图2和图3中的方法实施例中的相应流程,具体可参见上述方法实施例中的描述,为避免重复,此处适当省略详述描述。
具体的,该通信网络包括该控制器和多个节点,该多个节点包括灵活节点和非灵活节点,该灵活节点与该控制器直接连接,该非灵活节点与该控制器间接相连。图5所示的控制器500包括:处理器510和存储器520,可选地,该处理器500还可以包括总线系统530。其中,处理器510和存储器520通过总线系统530相连。该存储器520用于存储指令,该处理器510用于执行该存储器520存储的指令从该灵活节点中确定至少一个采样节点和流量采样的时间频率,通知该至少一个采样节点以该时间频率进行流量采样,并接收每一次采样的所有采样节点的流量采样数据;根据每一次采样的该流量采样数据和该通信网络中的链路利用率高于第一预设值或低于第二预设值的链路的总流量信息确定流量工程所需要的参数信息;根据该参数信息进行流量工程控制。
因此,本发明实施例通过确定出的至少一个采样节点以时间频率进行流量采样,该控制器通知该至少一个采样节点以时间频率采样数据,并接收每一次采样的所有采样节点的流量采样数据;并根据该流量采用数据确定流量工程所需要的参数信息,最后根据该参数信息进行流量工程控制。实现了在有限采样下的TE控制。
上述本发明实施例揭示的方法可以应用于处理器510中,或者由处理器510实现。处理器510可能是一种集成电路芯片,具有信号的处理能力。在实现过程中,上述方法的各步骤可以通过处理器510中的硬件的集成逻辑电路或者软件形式的指令完成。上述的处理器510可以是通用处理器、数字信号处理器(Digital Signal Processor,DSP)、专用集成电路(Application Specific Integrated Circuit,ASIC)、现场可编程门阵列(Field Programmable Gate Array,FPGA)或者其他可编程逻辑器件、分立门或者晶体管逻辑器件、分立硬件组件。可以实现或者执行本发明实施例中的公开的各方法、步骤及逻辑框图。通用处理器可以是微处理器或者该处理器也可以是任何常规的处理器等。结合本发明实施例所公开的方法的步骤可以直接体现为硬件译码处理器执行完成,或者用译码处理器中的硬件及软件模块组合执行完成。软件模块可以位于随机存取存储器(Random Access Memory,RAM)、闪存、只读存储器(Read-Only Memory,ROM)、可编程只读存储器或者电可擦写可编程存储器、寄存器等本领域成熟的存储介质中。该存储介质位于存储器520,处理器510读取存储器520中的信息,结合其硬件完成上述方法的步骤,该总线系统530除包括数据总线之外,还可以包括电源总线、控制总线和状态信号总线等。但是为了清楚说明起见,在图中将各种总线都标为总线系统530。
可选地,作为另一实施例,该流量采样数据包括该通信网络中经过或起始于每一个采样节点到该通信网络中的每一个目的节点的业务流量。
可选地,作为另一实施例,处理器510具体用于根据该通信网络中业务的预估总流量确定该至少一个采样节点和流量采样的时间频率。
进一步地,作为另一实施例,处理器510具体用于根据以下公式确定该至少一个采样节点的个数和该时间频率:
|Cs|=γ11|C|
V*Ts=γ12log(|Cs|)
或者
|Cs|=log(γ21|C|)
V*Ts=γ22|Cs|,
其中,|Cs|表示该采样节点的个数,Ts表示该采样频率,V表示该预估总流量,C表示该通信网络中灵活节点的集合,|C|表示该通信网络中灵活节点的个数,γ11,γ12,γ21,γ22为预设常数参数,且0<γ11<1,0<log(γ21|C|)<|C|,
在确定该采样节点个数后,处理器510在该灵活节点集合中随机选取|Cs|个节点作为该采样节点的集合;
或者,处理器510具体用于对每一个目的节点d,根据以下公式确定该至少一个采样节点的个数和该时间频率:
|Cs
d|=γ11
d|C|
Vd*Ts
d=γ12
dlog(|Cs
d|)
或者
|Cs
d|=log(γ21
d|C|)
Vd*Ts
d=γ22
d|Cs
d|,
其中,|Cs
d|表示对应该目的节点d的采样节点的个数,Ts
d表示对应该目的节点d的采样频率,Vd表示到该目的节点d的预估总流量,C表示该通信网络中灵活节点的集合,|C|表示该d通信网络中灵活节点的个数,γ11
d,γ12
d,γ21
d,γ22
d为预设常数参数,且0<γ11
d<1,0<log(γ21
d|C|)<|C|。
在确定该采样节点个数后,处理器510针对每一个目的节点d,在预先基于灵活节点的拓扑生成的有向非循环图DAG中的路径中选取的|Cs
d|个节点,作为该采样节点的集合,或在所述灵活节点集合中随机选取|Cs
d|个节点,作为所述采样节点的集合。
可选地,作为另一实施例,该流量工程所需要的参数信息包括每一个灵活节点到该通信网络中的每一个目的节点的注入业务流量和该通信网络中的链路利用率高于预设值的链路中不可控的业务量,
该处理器510具体用于根据公式(1)确定流量工程所需要的参数信息。
可替代地,作为另一实施例,该流量工程所需要的参数信息包括每一个灵活节点到该通信网络中的每一个目的节点的注入业务流量和该通信网络中的链路利用率高于预设值的链路中不可控的业务量,
该处理器510具体用于根据公式(2)、公式(3)和公式(4)中的任意一种确定流量工程所需要的参数信息。
可选地,作为另一实施例,该处理器510具体用于根据该参数信息最小化所述通信网络中的所有路径的成本。
进一步地,作为另一实施例,该处理器510具体用于根据公式(5)或(6)最小化所述通信网络中的所有路径的成本。
可替代的,作为另一实施例,该控制单元可以根据(5)的特殊例子,如公式(5-11)、公式(5-12)、公式(5-13)、公式(5-14)和公式(5-15);和公式(5-21)、公式(5-22)、公式(5-23)、公式(5-24)和公式(5-25);以及公式(6)的特殊例子公式(6-11)和公式(6-12);以及公式(6-21)和公式(6-22)中任一种公式最小化路径成本。具体
地,可以参照对应方法实施例的各个过程,为避免重复,此处不再赘述。
因此,通过上述方案,本发明实例,能够实现快速TE控制,并且能够以低复杂度的算法去实现不同现有技术认为高复杂度的现有TE控制问题。换句话说,本发明实施例提出并解决更普遍性的TE控制问题,使实时TE可以实现。
应理解,说明书通篇中提到的“一个实施例”或“一实施例”意味着与实施例有关的特定特征、结构或特性包括在本发明的至少一个实施例中。因此,在整个说明书各处出现的“在一个实施例中”或“在一实施例中”未必一定指相同的实施例。此外,这些特定的特征、结构或特性可以任意适合的方式结合在一个或多个实施例中。应理解,在本发明的各种实施例中,上述各过程的序号的大小并不意味着执行顺序的先后,各过程的执行顺序应以其功能和内在逻辑确定,而不应对本发明实施例的实施过程构成任何限定。
另外,本文中术语“系统”和“网络”在本文中常被可互换使用。本文中术语“和/或”,仅仅是一种描述关联对象的关联关系,表示可以存在三种关系,例如,A和/或B,可以表示:单独存在A,同时存在A和B,单独存在B这三种情况。另外,本文中字符“/”,一般表示前后关联对象是一种“或”的关系。
应理解,在本发明实施例中,“与A相应的B”表示B与A相关联,根据A可以确定B。但还应理解,根据A确定B并不意味着仅仅根据A确定B,还可以根据A和/或其它信息确定B。
本领域普通技术人员可以意识到,结合本文中所公开的实施例描述的各示例的单元及算法步骤,能够以电子硬件、计算机软件或者二者的结合来实现,为了清楚地说明硬件和软件的可互换性,在上述说明中已经按照功能一般性地描述了各示例的组成及步骤。这些功能究竟以硬件还是软件方式来执行,取决于技术方案的特定应用和设计约束条件。专业技术人员可以对每个特定的应用来使用不同方法来实现所描述的功能,但是这种实现不应认为超出本发明的范围。
所属领域的技术人员可以清楚地了解到,为了描述的方便和简洁,上述描述的系统、装置和单元的具体工作过程,可以参考前述方法实施例中的对应过程,在此不再赘述。
在本申请所提供的几个实施例中,应该理解到,所揭露的系统、装置和方法,可以通过其它的方式实现。例如,以上所描述的装置实施例仅仅是示意性的,例如,单元的划分,仅仅为一种逻辑功能划分,实际实现时可以有另外的划分方式,例如多个单元或组件可以结合或者可以集成到另一个系统,或一些特征可以忽略,或不执行。另外,所显示或讨论的相互之间的耦合或直接耦合或通信连接可以是通过一些接口、装置或单元的间接耦合或通信连接,也可以是电的,机械的或其它的形式连接。
作为分离部件说明的单元可以是或者也可以不是物理上分开的,作为单元显示的部件可以是或者也可以不是物理单元,即可以位于一个地方,或者也可以分布到多个网络单元上。可以根据实际的需要选择其中的部分或者全部单元来实现本发明实施例方案的目的。
另外,在本发明各个实施例中的各功能单元可以集成在一个处理单元中,也可以是各个单元单独物理存在,也可以是两个或两个以上单元集成在一个单元中。上述集成的单元既可以采用硬件的形式实现,也可以采用软件功能单元的形式实现。
通过以上的实施方式的描述,所属领域的技术人员可以清楚地了解到本发明实施例
可以用硬件实现,或固件实现,或它们的组合方式来实现。当使用软件实现时,可以将上述功能存储在计算机可读介质中或作为计算机可读介质上的一个或多个指令或代码进行传输。计算机可读介质包括计算机存储介质和通信介质,其中通信介质包括便于从一个地方向另一个地方传送计算机程序的任何介质。存储介质可以是计算机能够存取的任何可用介质。以此为例但不限于:计算机可读介质可以包括RAM、ROM、EEPROM、CD-ROM或其他光盘存储、磁盘存储介质或者其他磁存储设备、或者能够用于携带或存储具有指令或数据结构形式的期望的程序代码并能够由计算机存取的任何其他介质。此外。任何连接可以适当的成为计算机可读介质。例如,如果软件是使用同轴电缆、光纤光缆、双绞线、数字用户线(DSL)或者诸如红外线、无线电和微波之类的无线技术从网站、服务器或者其他远程源传输的,那么同轴电缆、光纤光缆、双绞线、DSL或者诸如红外线、无线和微波之类的无线技术包括在所属介质的定影中。如本发明实施例所使用的,盘(Disk)和碟(disc)包括压缩光碟(CD)、激光碟、光碟、数字通用光碟(DVD)、软盘和蓝光光碟,其中盘通常磁性的复制数据,而碟则用激光来光学的复制数据。上面的组合也应当包括在计算机可读介质的保护范围之内。
总之,以上所述仅为本发明实施例技术方案的较佳实施例而已,并非用于限定本发明实施例的保护范围。凡在本发明实施例的精神和原则之内,所作的任何修改、等同替换、改进等,均应包含在本发明实施例的保护范围之内。
Claims (38)
- 一种用于通信网络中的流量工程的方法,其特征在于,所述通信网络包括控制器和多个节点,所述多个节点包括灵活节点和非灵活节点,所述灵活节点与所述控制器直接连接,所述非灵活节点与所述控制器间接相连,所述方法包括:所述控制器从所述灵活节点中确定至少一个采样节点和流量采样的时间频率,所述控制器通知所述至少一个采样节点以所述时间频率进行流量采样,并接收每一次采样的所有采样节点的流量采样数据;所述控制器根据每一次采样的所述流量采样数据和所述通信网络中的链路利用率高于第一预设值或低于第二预设值的链路的总流量信息确定流量工程所需要的参数信息;所述控制器根据所述参数信息进行流量工程控制。
- 根据权利要求1所述的方法,其特征在于,所述流量采样数据包括所述通信网络中经过或起始于每一个采样节点到所述通信网络中的每一个目的节点的业务流量。
- 根据权利要求1或2所述的方法,其特征在于,所述控制器从所述灵活节点中确定至少一个采样节点和流量采样的时间频率,包括:所述控制器根据所述通信网络中业务的预估总流量确定所述至少一个采样节点和流量采样的时间频率。
- 根据权利要求3所述的方法,其特征在于,所述控制器根据所述通信网络中业务的预估总流量确定所述至少一个采样节点和流量采样的时间频率,包括:所述控制器根据以下公式确定所述至少一个采样节点的个数和所述时间频率:|Cs|=γ11|C|V*Ts=γ12log(|Cs|)或者|Cs|=log(γ21|C|)V*Ts=γ22|Cs|,其中,|Cs|表示所述采样节点的个数,Ts表示所述采样频率,V表示所述预估总流量,C表示所述通信网络中灵活节点的集合,|C|表示所述通信网络中灵活节点的个数,γ11,γ12,γ21,γ22为预设常数参数,且0<γ11<1,0<log(γ21|C|)<|C|,在确定所述采样节点个数后,所述控制器在所述灵活节点集合中随机选取|Cs|个节点,作为所述采样节点的集合;或者,所述控制器,对每一个目的节点d,根据以下公式确定所述至少一个采样节点的个数和所述时间频率:或者|Cs d|=log(γ21 d|C|)Vd*Ts d=γ22 d|Cs d|,其中,|Cs d|表示对应所述目的节点d的采样节点的个数,Ts d表示对应所述目的节点d的采样频率,Vd表示到所述目的节点d的预估总流量,C表示所述通信网络中灵活节点的集合,|C|表示所述通信网络中灵活节点的个数,γ11 d,γ12 d,γ21 d,γ22 d为预设常数参数,且0<γ11 d<1,0<log(γ21 d|C|)<|C|,在确定所述采样节点个数后,所述控制器针对每一个目的节点d,在预先基于灵活节点的拓扑生成的有向非循环图DAG中的路径中选取的|Cs d|个节点,作为所述采样节点的集合,或在所述灵活节点集合中随机选取|Cs d|个节点,作为所述采样节点的集合。
- 根据权利要求1至4中任一项所述的方法,其特征在于,所述流量工程所需要的参数信息包括每一个灵活节点到所述通信网络中的每一个目的节点的注入业务流量和所述通信网络中的链路利用率高于预设值的链路中不可控的业务量,所述控制器根据每一次采样的所述流量采样数据和所述通信网络中的链路利用率高于第一预设值或低于第二预设值的链路的总流量信息确定流量工程所需要的参数信息,包括根据以下公式(1)确定流量工程所需要的参数信息:限制条件为:Icurrent=Iud=ΔI+Ilast iteration其中,E表示所述通信网络中所有链路集合,N表示所述通信网络中所有节点的集合,C表示所述通信网络中所有灵活节点的集合;其中,w表示采样节点,u表示灵活节点,d表示目的节点,ΔI是指上次采样时灵活节点u的注入流量Ilast iteration与当前采样时灵活节点u的注入流量Iud的差,vec(ΔI)表示把所有元件ΔI(对于所有的u)组成向量,表示vec(ΔI)的最小l0范数,e表示所述通信网络中的链路利用率高于第一预设值的一个链路或低于第二预设值的一个链路,f(e)表示链路e的总流量,g(e)表示链路e中不可控的业务量,Wwd表示所述通信网络中经过或起始于采样节点w到所述通信网络中的节点d的业务流量,Iud表示灵活节点u到目的节点d的注入业务流量,从节点u发出一个包去目的地d,经过链路e的总份额表示为αe(u,d),经过采样节点w的总份额表示为βw(u,d),且当w被u发出的包经过表示成u≤w。
- 根据权利要求1至4中任一项所述的方法,其特征在于,所述流量工程所需要的参数信息包括每一个灵活节点到所述通信网络中的每一个目的节点的注入业务流量和所述通信网络中的链路利用率高于预设值的链路中不可控的业务量,所述控制器根据每一次采样的所述流量采样数据和所述通信网络中的链路利用率高于第一预设值或低于第二预设值的链路的总流量信息确定流量工程所需要的参数信息,包括根据以下公式(2)、公式(3)和公式(4)中的任意一个确定流量工程所需要的参数信息:限制条件为:Icurrent=Iud=ΔI+Ilast iteration其中,E表示所述通信网络中所有链路集合,N表示所述通信网络中所有节点的集合,C表示所述通信网络中所有灵活节点的集合;其中,w表示采样节点,u表示灵活节点,d表示目的节点,ΔI是指上次采样时灵活节点u的注入流量Ilast iteration与当前采样时灵活节点u的注入流量Iud的差,vec(ΔI)表示把所有元件ΔI(对于所有的u)组成向量,表示vec(ΔI)的最小l1范数,e表示所述通信网络中的链路利用率高于第一预设值的一个链路或低于第二预设值的一个链路,f(e)表示链路e的总流量,g(e)表示链路e中不可控的业务量,Wwd表示所述通信网络中经过或起始于采样节点w到所述通信网络中的节点d的业务流量,Iud表示灵活节点u到目的节点d的注入业务流量,从节点u发出一个包去目的地d,经过链路e的总份额表示为αe(u,d),经过采样节点w的总份额表示为βw(u,d),且当w被u发出的包经过表示成u≤w;限制条件为:Icurrent=Iud=ΔI+Ilast iteration限制条件为:Icurrent=Iud=ΔI+Ilast iteration
- 根据权利要求1至6中任一项所述的方法,其特征在于,所述控制器根据所述参数信息进行流量工程控制,包括:所述控制器根据所述参数信息最小化所述通信网络中 的所有路径的成本。
- 根据权利要求7所述的方法,其特征在于,所述控制器根据所述参数信息最小化所述通信网络中所有路径的成本,包括所述控制器根据以下公式(5)最小化所述通信网络的网络成本:限制条件为:
- 根据权利要求7所述的方法,其特征在于,所述控制器根据所述参数信息最小化所述通信网络中的所有路径的成本,包括所述控制器根据以下公式(6)最小化所述通信网络中的所有路径的最大成本:限制条件为:
- 根据权利要求7所述的方法,其特征在于,所述控制器根据所述参数信息最小化所述通信网络中所有路径的成本,包括所述控制器根据公式(5-11)最小化所述通信网络中的网络成本:限制条件为:0≤he(θ)≤1其中,e表示所述链路,E表示所有链路的集合,C表示灵活节点的集合,u表示灵活节点,d表示目的节点,N表示所有节点的集合,Iud表示灵活节点u到目的节点d的注入业务流量,Pud表示所有从u至d的可行路径的集合,P∈Pud表示P表示Pud中的一条可行路径,x(P)表示路径P的流量,表示所有从u至d的可行路径的总流量,λ表示最大的网络吞吐量比例,θ表示链路的最大利用率相关参数,he(θ)用来表示权重参数与θ的关系函数,其中权重参数包括γe和γP,γe表示链路e的权重参数,γP路径P的权重参数,c(e)表示链路e的总带宽,g(e)表示链路e的不可控流的流量,表示链路e上的所有可行路径的可控流的总流量,其中,根据权重参数γe和γP的不同权重情况,he(θ)可以为以下中的任意一种:
- 根据权利要求7所述的方法,其特征在于,所述控制器根据所述参数信息最小化所述通信网络中所有路径的成本,包括所述控制器根据公式(5-21)最小化所述通信网络中的网络成本:限制条件为:0≤he(θ)≤1其中,e表示所述链路,E表示所有链路的集合,C表示灵活节点的集合,u表示灵活节点,d表示目的节点,N表示所有节点的集合,Iud表示灵活节点u到目的节点d的注入业务流量,Pud表示所有从u至d的可行路径的集合,P∈Pud表示P表示Pud中的一条可行路径,x(P)表示路径P的流量,λ表示最大的网络吞吐量比例,表示所有从u至d的可行路径的总流量,θ表示链路的均衡相关参数,he(θ)用来表示权重参数与θ的关系函数,其中权重参数包括γe和γP,γe表示链路e的权重参数,γP路径P的权重参数,c(e)表示链路e的总带宽,g(e)表示链路e的不可控流的流量,表示链路e上的所有可行路径的可控流的总流量,其中,根据权重参数γe和γP的不同权重情况,he(θ)可以为以下中的任意一种:
- 根据权利要求7所述的方法,其特征在于,所述控制器根据所述参数信息最小化所述通信网络中所有路径的成本,包括所述控制器根据公式(6-11)最小化所述通信网络中的网络成本:限制条件为:0≤ke(λ)≤λ
- 根据权利要求12所述的方法,其特征在于,所述控制器根式(6-11)最小化所述通信网络中的网络成本,包括:所述控制器将所述ke(λ)作为常数,并采用低复杂度的最大并发分数流算法求解公式(6-11)最小化所述通信网络中的网络成本。
- 根据权利要求7所述的方法,其特征在于,所述控制器根据所述参数信息最小化所述通信网络中所有路径的成本,包括所述控制器根据公式(6-12)最小化所述通信网络中的所有路径的最大成本:限制条件为:0≤ke(λ)≤λ其中,e表示所述链路,E表示所有链路的集合,C表示灵活节点的集合,u表示灵活节点,d表示目的节点,N表示所有节点的集合,Iud表示灵活节点u到目的节点d的注入业务流量,Pud表示所有从u至d的可行路径的集合,P∈Pud表示P表示Pud中的一条可行路径,x(P)表示路径P的流量,表示所有从u至d的可行路径的总流量,λ表示最大的网络吞吐量相关参数,ke(λ)用于表示链路e的权重参数与λ的对应关系,其中γe表示链路的权重参数,ρ(e)表示链路e的可控带宽的利用率,c(e)表示链路e的总带宽,g(e)表示链路e的不可控流的流量,表示链路e上的所有可行路径的可控流的总流量,其中,根据权重参数γe不同权重情况,ke(λ)可以为以下中的任意一种:
- 根据权利要求14所述的方法,其特征在于,所述控制器以下公式:公式(6-12)最小化所述通信网络中的网络成本,包括:所述控制将ke(λ)作为常数,并采用低复杂度的最大并发分数流算法求解公式(6-12)最小化所述通信网络中的所有路径的最大成本。
- 根据权利要求7所述的方法,其特征在于,所述控制器根据所述参数信息最小化所述通信网络中所有路径的成本,包括所述控制器根据公式(6-21)最小化所述通信网络中的网络成本:限制条件为:0≤ke(λ)≤1其中,e表示所述链路,E表示所有链路的集合,C表示灵活节点的集合,u表示灵活节点,d表示目的节点,N表示所有节点的集合,Iud表示灵活节点u到目的节点d的注入业务流量,Pud表示所有从u至d的可行路径的集合,P∈Pud表示P表示Pud中的一条可行路径,x(P)表示路径P的流量,λ表示均衡网络下的网络吞吐量相关参数,表示所有从u至d的可行路径的总流量,ke(λ)用于表示链路e的权重参数与λ的对应关系,其中权重参数包括γe和γP,γe表示链路权重参数,γP表示路径权重参数,ρ(e)表示链路e的可控带宽的利用率,c(e)表示链路e的总带宽,g(e)表示链路e的不可控流的流量,表示链路e上的所有可行路径的可控流的总流量。
- 根据权利要求16所述的方法,其特征在于,所述控制器根据公式(6-21)最小化所述通信网络中的网络成本,包括:所述控制器将所述ke(λ)作为常数,并采用改进的最大并发分数流算法求解公式(6-21)最小化所述通信网络中的网络成本。
- 根据权利要求7所述的方法,其特征在于,所述控制器根据所述参数信息最小化所述通信网络中所有路径的成本,包括所述控制器根据公式(6-22)最小化所述通信 网络中的所有路径的最大成本:限制条件为:0≤ke(λ)≤1其中,e表示所述链路,E表示所有链路的集合,C表示灵活节点的集合,u表示灵活节点,d表示目的节点,N表示所有节点的集合,Iud表示灵活节点u到目的节点d的注入业务流量,Pud表示所有从u至d的可行路径的集合,P∈Pud表示P表示Pud中的一条可行路径,x(P)表示路径P的流量,λ表示均衡网络下的网络吞吐量相关参数,表示所有从u至d的可行路径的总流量,ke(λ)用于表示链路e的权重参数与λ的对应关系,其中权重参数包括γe和γP,γe表示链路权重参数,γP表示路径权重参数,ρ(e)表示链路e的可控带宽的利用率,c(e)表示链路e的总带宽,g(e)表示链路e的不可控流的流量,表示链路e上的所有可行路径的可控流的总流量;其中,根据权重参数γe不同权重情况,ke(λ)可以为以下中的任意一种:ke(λ)=γeλ;
- 根据权利要求18所述的方法,其特征在于,所述控制器根据公式(6-22)最小化所述通信网络中的网络成本,包括:所述控制将ke(λ)作为常数,并采用改进的最大并发分数流算法求解公式(6-22)最小化所述通信网络中的所有路径的最大成本。
- 一种通信网络中的控制器,其特征在于,所述通信网络包括所述控制器和多个节点,所述多个节点包括灵活节点和非灵活节点,所述灵活节点与所述控制器直接连接,所述非灵活节点与所述控制器间接相连,所述控制器包括:控制单元,用于从所述灵活节点中确定至少一个采样节点和流量采样的时间频率,收发单元,用于通知所述至少一个采样节点以所述时间频率进行流量采样,并接收每一次采样的所有采样节点的流量采样数据;确定单元,用于根据每一次采样的所述流量采样数据和所述通信网络中的链路利用率高于第一预设值或低于第二预设值的链路的总流量信息确定流量工程所需要的参数信息;控制单元,用于根据所述参数信息进行流量工程控制。
- 根据权利要求20所述的控制器,其特征在于,所述流量采样数据包括所述通信网络中经过或起始于每一个采样节点到所述通信网络中的每一个目的节点的业务流量。
- 根据权利要求20或21所述的控制器,其特征在于,所述控制单元具体用于根据所述通信网络中业务的预估总流量确定所述至少一个采样节点和流量采样的时间频率。
- 根据权利要求22所述的控制器,其特征在于,所述控制单元具体用于根据以下公式确定所述至少一个采样节点的个数和所述时间频率:|Cs|=γ11|C|V*Ts=γ12log(|Cs|)或者|Cs|=log(γ21|C|)V*Ts=γ22|Cs|,其中,|Cs|表示所述采样节点的个数,Ts表示所述采样频率,V表示所述预估总流量,C表示所述通信网络中灵活节点的集合,|C|表示所述通信网络中灵活节点的个数,γ11,γ12,γ21,γ22为预设常数参数,且0<γ11<1,0<log(γ21|C|)<|C|,在确定所述采样节点个数后,所述控制器在所述灵活节点集合中随机选取|Cs|个节点作为所述采样节点的集合;或者,所述控制单元具体用于对每一个目的节点d,根据以下公式确定所述至少一个采样节点的个数和所述时间频率:或者|Cs d|=log(γ21 d|C|)Vd*Ts d=γ22 d|Cs d|,其中,|Cs d|表示对应所述目的节点d的采样节点的个数,Ts d表示对应所述目的节点d的采样频率,Vd表示到所述目的节点d的预估总流量,C表示所述通信网络中灵活节点的集合,|C|表示所述通信网络中灵活节点的个数,γ11 d,γ12 d,γ21 d,γ22 d为预设常数参数,且0<γ11 d<1,0<log(γ21 d|C|)<|C|,在确定所述采样节点个数后,所述控制单元针对每一个目的节点d,在预先基于灵活节点的拓扑生成的有向非循环图DAG中的路径中选取的|Cs d|个节点,作为所述采样节点的集合,或在所述灵活节点集合中随机选取|Cs d|个节点,作为所述采样节点的集合。
- 根据权利要求20至23中任一项所述的控制器,其特征在于,所述流量工程所需要的参数信息包括每一个灵活节点到所述通信网络中的每一个目的节点的注入业务流量和所述通信网络中的链路利用率高于预设值的链路中不可控的业务量,所述确定单元具体用于根据以下公式(1)确定流量工程所需要的参数信息:限制条件为:Icurrent=Iud=ΔI+Ilast iteration其中,E表示所述通信网络中所有链路集合,N表示所述通信网络中所有灵活节点的集合,C表示所述通信网络中所有灵活节点的集合;其中,w表示采样节点,u表示灵活节点,d表示目的节点,ΔI是指上次采样时灵活节点u的注入流量Ilast iteration与当前采样时灵活节点u的注入流量Iud的差,vec(ΔI)表示把所有元件ΔI(对于所有的u)组成向量,表示vec(ΔI)的最小l0范数,e表示所述通信网络中的链路利用率高于第一预设值的一个链路或低于第二预设值的一个链路,f(e)表示链路e的总流量,g(e)表示链路e中不可控的业务量,Wwd表示所述通信网络中经过或起始于采样节点w到所述通信网络中的节点d的业务流量,Iud表示灵活节点u到目的节点d的注入业务流量,从节点u发出一个包去目的地d,经过链路e的总份额表示为αe(u,d),经过采样节点w的总份额表示为βw(u,d),且当w被u发出的包经过表示成u≤w。
- 根据权利要求20至23中任一项所述的控制器,其特征在于,所述流量工程所需要的参数信息包括每一个灵活节点到所述通信网络中的每一个目的节点的注入业务流量和所述通信网络中的链路利用率高于预设值的链路中不可控的业务量,所述确定单元具体用于根据以下公式(2)、公式(3)和公式(4)中的任意一种确定流量工程所需要的参数信息:限制条件为:Icurrent=Iud=ΔI+Ilast iteration其中,E表示所述通信网络中所有链路集合,N表示所述通信网络中所有灵活节点的集合,C表示所述通信网络中所有灵活节点的集合;其中,w表示采样节点,u表示灵活节点,d表示目的节点,ΔI是指上次采样时灵活节点u的注入流量Ilast iteration与当前采样时灵活节点u的注入流量Iud的差,vec(ΔI)表示把所有元件ΔI(对于所有的u)组成向量,表示vec(ΔI)的最小l1范数,e表示所述通信网络中的链路利用率高于第一预设值的一个链路或低于第二预设值的一个链路,f(e)表示链路e的总流量,g(e)表示链路e中不可控的业务量,Wwd表示所述通信网络中经过或起始于采样节点w到所述通信网络中的节点d的业务流量,Iud表示灵活节点u到目的节点d的注入业务流 量,从节点u发出一个包去目的地d,经过链路e的总份额表示为αe(u,d),经过采样节点w的总份额表示为βw(u,d),且当w被u发出的包经过表示成u≤w;限制条件为:Icurrent=Iud=ΔI+Ilast iteration限制条件为:Icurrent=Iud=ΔI+Ilast iteration
- 根据权利要求20至25中任一项所述的控制器,其特征在于,所述控制单元具体用于根据所述参数信息最小化所述通信网络中所有路径的成本。
- 根据权利要求26所述的控制器,其特征在于,所述控制单元具体用于根据以下公式(5)最小化所述通信网络的网络成本:限制条件为:
- 根据权利要求26所述的控制器,其特征在于,所述控制单元具体用于根据以下公式(6)最小化所述通信网络中的所有路径的成本:限制条件为:
- 根据权利要求26所述的控制器,其特征在于,所述控制单元具体用于根据公式(5-11)最小化所述通信网络中的所有路径的成本:限制条件为:0≤he(θ)≤1其中,e表示所述链路,E表示所有链路的集合,C表示灵活节点的集合,u表示灵活节点,d表示目的节点,N表示所有节点的集合,Iud表示灵活节点u到目的节点d的注入业务流量,Pud表示所有从u至d的可行路径的集合,P∈Pud表示P表示Pud中的一条可行路径,x(P)表示路径P的流量,表示所有从u至d的可行路径的总流量, λ表示最大的网络吞吐量比例,θ表示链路的最大利用率相关参数,he(θ)用来表示权重参数与θ的关系函数,其中权重参数包括γe和γP,γe表示链路e的权重参数,γP路径P的权重参数),c(e)表示链路e的总带宽,g(e)表示链路e的不可控流的流量,表示链路e上的所有可行路径的可控流的总流量,其中,根据权重参数γe和γP的不同权重情况,he(θ)可以为以下中的任意一种:
- 根据权利要求26所述的控制器,其特征在于,所述控制单元具体用于根据公式(5-21)最小化所述通信网络中的网络成本:限制条件为:0≤he(θ)≤1其中,e表示所述链路,E表示所有链路的集合,C表示灵活节点的集合,u表示灵活节点,d表示目的节点,N表示所有节点的集合,Iud表示灵活节点u到目的节点d的注入业务流量,Pud表示所有从u至d的可行路径的集合,P∈Pud表示P表示Pud中的一条可行路径,x(P)表示路径P的流量,λ表示最大的网络吞吐量比例,表示所有从u至d的可行路径的总流量,θ表示链路的均衡相关参数,he(θ)用来表示权重参数与θ的关系函数,其中权重参数包括γe和γP,γe表示链路e的权重参数,γP路径P的权重参数),c(e)表示链路e的总带宽,g(e)表示链路e的不可控流的流量,表示链路e上的所有可行路径的可控流的总流量,其中,根据权重参数γe和γP的不同权重情况,he(θ)可以为以下中的任意一种:
- 根据权利要求26所述的控制器,其特征在于,所述控制单元具体用于根据公式(6-11)最小化所述通信网络中的网络成本:限制条件为:0≤ke(λ)≤λ
- 根据权利要求31所述的控制器,其特征在于,所述控制单元具体用于将所述ke(λ)作为常数,并采用低复杂度的最大并发分数流算法求解公式(6-11)最小化所述通信网络中的网络成本。
- 根据权利要求26所述的控制器,其特征在于,所述控制单元具体用于根据公式(6-12)最小化所述通信网络中的所有路径的最大成本:限制条件为:0≤ke(λ)≤λ其中,e表示所述链路,E表示所有链路的集合,C表示灵活节点的集合,u表示灵活节点,d表示目的节点,N表示所有节点的集合,Iud表示灵活节点u到目的节点d的注入业务流量,Pud表示所有从u至d的可行路径的集合,P∈Pud表示P表示Pud中的一条可行路径,x(P)表示路径P的流量,表示所有从u至d的可行路径的总流量,λ表示最大的网络吞吐量相关参数,ke(λ)用于表示链路e的权重参数与λ的对应关系,其中γe表示链路的权重参数,ρ(e)表示链路e的可控带宽的利用率,c(e)表示链路e的总带宽,g(e)表示链路e的不可控流的流量,表示链路e上的所有可行路径的可控流的总流量,其中,根据权重参数γe不同权重情况,ke(λ)可以为以下中的任意一种:
- 根据权利要求33所述的控制器,其特征在于,所述控制单元具体用于:将ke(λ)作为常数,并采用低复杂度的最大并发分数流算法求解公式(6-12)最小化所述通信网络中的所有路径的最大成本。
- 根据权利要求26所述的控制器,其特征在于,所述控制单元具体用于根据公式(6-21)最小化所述通信网络中的网络成本:限制条件为:0≤ke(λ)≤1其中,e表示所述链路,E表示所有链路的集合,C表示灵活节点的集合,u表示灵活节点,d表示目的节点,N表示所有节点的集合,Iud表示灵活节点u到目的节点d的 注入业务流量,Pud表示所有从u至d的可行路径的集合,P∈Pud表示P表示Pud中的一条可行路径,x(P)表示路径P的流量,λ表示均衡网络下的网络吞吐量相关参数,表示所有从u至d的可行路径的总流量,ke(λ)用于表示链路e的权重参数与λ的对应关系,其中权重参数包括γe和γP,γe表示链路权重参数,γP表示路径权重参数,ρ(e)表示链路e的可控带宽的利用率,c(e)表示链路e的总带宽,g(e)表示链路e的不可控流的流量,表示链路e上的所有可行路径的可控流的总流量。
- 根据权利要求35所述的控制器,其特征在于,所述控制单元具体用于将所述ke(λ)作为常数,并采用改进的最大并发分数流算法(Fractional Maximum Concurrent Flow)求解公式(6-21)最小化所述通信网络中的网络成本。
- 根据权利要求26所述的控制器,其特征在于,所述控制单元具体用于根据公式(6-22)最小化所述通信网络中的所有路径的最大成本:限制条件为:0≤ke(λ)≤1其中,e表示所述链路,E表示所有链路的集合,C表示灵活节点的集合,u表示灵活节点,d表示目的节点,N表示所有节点的集合,Iud表示灵活节点u到目的节点d的注入业务流量,Pud表示所有从u至d的可行路径的集合,P∈Pud表示P表示Pud中的一条可行路径,x(P)表示路径P的流量,λ表示均衡网络下的网络吞吐量相关参数,表示所有从u至d的可行路径的总流量,ke(λ)用于表示链路e的权重参数与λ的对应关系,其中权重参数包括γe和γP,γe表示链路权重参数,γP表示路径权重参数,ρ(e)表示链路e的可控带宽的利用率,c(e)表示链路e的总带宽,g(e)表示链路e的不可控流的流量,表示链路e上的所有可行路径的可控流的总流量;其中,根据权重参数γe不同权重情况,ke(λ)可以为以下中的任意一种:ke(λ)=γeλ;
- 根据权利要求37所述的控制器,其特征在于,所述控制单元具体用于:将ke(λ)作为常数,并采用改进的最大并发分数流算法求解公式(6-22)最小化所述通信网络中的所有路径的最大成本。
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| EP17746907.9A EP3399706B1 (en) | 2016-02-05 | 2017-01-24 | Traffic engineering method for use in communication network, and controller |
| US16/054,087 US10764161B2 (en) | 2016-02-05 | 2018-08-03 | Method for traffic engineering in communications network and controller |
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610081473.8 | 2016-02-05 | ||
| CN201610081473.8A CN107046504B (zh) | 2016-02-05 | 2016-02-05 | 用于通信网络中的流量工程的方法和控制器 |
Related Child Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| US16/054,087 Continuation US10764161B2 (en) | 2016-02-05 | 2018-08-03 | Method for traffic engineering in communications network and controller |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2017133587A1 true WO2017133587A1 (zh) | 2017-08-10 |
Family
ID=59499357
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/CN2017/072446 Ceased WO2017133587A1 (zh) | 2016-02-05 | 2017-01-24 | 用于通信网络中的流量工程的方法和控制器 |
Country Status (4)
| Country | Link |
|---|---|
| US (1) | US10764161B2 (zh) |
| EP (1) | EP3399706B1 (zh) |
| CN (1) | CN107046504B (zh) |
| WO (1) | WO2017133587A1 (zh) |
Families Citing this family (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109873762B (zh) * | 2017-12-05 | 2021-08-31 | 中国电信股份有限公司 | 路径调度方法、装置和计算机可读存储介质 |
| US11228603B1 (en) * | 2018-09-27 | 2022-01-18 | Juniper Networks, Inc. | Learning driven dynamic threat treatment for a software defined networking environment |
| CN110365545B (zh) * | 2019-08-09 | 2022-08-09 | 北京达佳互联信息技术有限公司 | 下发率处理方法、装置、服务器及存储介质 |
| US11088996B1 (en) * | 2021-02-10 | 2021-08-10 | SecureCo, Inc. | Secure network protocol and transit system to protect communications deliverability and attribution |
| CN113595756B (zh) * | 2021-06-11 | 2024-04-16 | 西安邮电大学 | 一种异质化节点和链路的网络建模方法、通信设备及网络 |
| US12347317B2 (en) * | 2022-03-29 | 2025-07-01 | Mitsubishi Electric Research Laboratories, Inc. | System and method for jointly controlling connected autonomous vehicles (CAVs) and manual connected vehicles (MCVs) |
| US12367771B2 (en) * | 2023-03-09 | 2025-07-22 | Mitsubishi Electric Research Laboratories, Inc. | Hierarchical optimization-based coordinated control of traffic rules and mixed traffic in multi-intersection environments |
| CN119966846A (zh) * | 2025-01-14 | 2025-05-09 | 上海勘测设计研究院有限公司 | 一种基于人工智能算法的流量监测方法及装置 |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101075911A (zh) * | 2006-05-18 | 2007-11-21 | 阿拉克斯拉网络株式会社 | 统计信息收集系统及统计信息收集装置 |
| CN102175269A (zh) * | 2011-01-24 | 2011-09-07 | 华东师范大学 | 一种可改变采样频率的传感器设备及其控制方法 |
| US20140112187A1 (en) * | 2012-10-23 | 2014-04-24 | Electronics And Telecommunications Research Institute | Apparatus for flow-based network monitoring and network monitoring system |
Family Cites Families (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8363744B2 (en) * | 2001-06-10 | 2013-01-29 | Aloft Media, Llc | Method and system for robust, secure, and high-efficiency voice and packet transmission over ad-hoc, mesh, and MIMO communication networks |
| CN100419444C (zh) * | 2002-10-18 | 2008-09-17 | 卡里德恩科技有限公司 | 用于在基于度量路由的网络中执行流量工程的方法和系统 |
| US7398438B2 (en) | 2006-03-30 | 2008-07-08 | Lucent Technologies Inc. | Method and apparatus for improved routing in connectionless networks |
| US20130272125A1 (en) * | 2010-12-22 | 2013-10-17 | Koninklijke Philips N.V. | Component, system and method for controlling communication of data of at least one application of a communications network |
| US8593958B2 (en) * | 2011-09-14 | 2013-11-26 | Telefonaktiebologet L M Ericsson (Publ) | Network-wide flow monitoring in split architecture networks |
| CN104579810B (zh) | 2013-10-23 | 2019-10-25 | 中兴通讯股份有限公司 | 软件定义网络流量采样方法和系统 |
| US9923832B2 (en) * | 2014-07-21 | 2018-03-20 | Cisco Technology, Inc. | Lightweight flow reporting in constrained networks |
| CN104796348B (zh) * | 2015-04-03 | 2018-02-13 | 华为技术有限公司 | 基于sdn的idc网络出口流量均衡调整方法、设备及系统 |
-
2016
- 2016-02-05 CN CN201610081473.8A patent/CN107046504B/zh active Active
-
2017
- 2017-01-24 WO PCT/CN2017/072446 patent/WO2017133587A1/zh not_active Ceased
- 2017-01-24 EP EP17746907.9A patent/EP3399706B1/en active Active
-
2018
- 2018-08-03 US US16/054,087 patent/US10764161B2/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101075911A (zh) * | 2006-05-18 | 2007-11-21 | 阿拉克斯拉网络株式会社 | 统计信息收集系统及统计信息收集装置 |
| CN102175269A (zh) * | 2011-01-24 | 2011-09-07 | 华东师范大学 | 一种可改变采样频率的传感器设备及其控制方法 |
| US20140112187A1 (en) * | 2012-10-23 | 2014-04-24 | Electronics And Telecommunications Research Institute | Apparatus for flow-based network monitoring and network monitoring system |
Non-Patent Citations (1)
| Title |
|---|
| See also references of EP3399706A4 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107046504A (zh) | 2017-08-15 |
| US10764161B2 (en) | 2020-09-01 |
| EP3399706B1 (en) | 2020-11-25 |
| CN107046504B (zh) | 2020-08-25 |
| EP3399706A1 (en) | 2018-11-07 |
| EP3399706A4 (en) | 2019-01-09 |
| US20180343176A1 (en) | 2018-11-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2017133587A1 (zh) | 用于通信网络中的流量工程的方法和控制器 | |
| CN111770028B (zh) | 用于计算机网络的方法和网络设备 | |
| US9559970B2 (en) | Shortening of service paths in service chains in a communications network | |
| US10897419B2 (en) | Method and apparatus for supporting service function chaining in a communication network | |
| EP2820808B1 (en) | Compound masking and entropy for data packet classification using tree-based binary pattern matching | |
| EP2926513B1 (en) | Packet prioritization in a software-defined network implementing openflow | |
| CN104702502B (zh) | 网络路径计算方法及装置 | |
| US8799507B2 (en) | Longest prefix match searches with variable numbers of prefixes | |
| WO2019011338A1 (zh) | 一种最短路径确定方法及控制器 | |
| JP6323547B2 (ja) | 通信システム、制御装置、通信制御方法、および、プログラム | |
| WO2022166348A1 (zh) | 路由方法、路由装置、控制器和计算机可读存储介质 | |
| EP3767886A1 (en) | Cluster oriented dynamic routing | |
| Alvarez-Horcajo et al. | A hybrid SDN switch based on standard P4 code | |
| Wang et al. | Low-latency service chaining with predefined NSH-based multipath across multiple datacenters | |
| KR101746105B1 (ko) | 서비스 체이닝이 가능한 오픈플로우 스위치 | |
| Nayyer et al. | Learning-based hybrid routing for scalability in software defined networks | |
| KR101679224B1 (ko) | Sdn 기반의 트래픽 분배 가능한 네트워크 시스템 | |
| Saha et al. | SAFAR: Simulated annealing-based flow allocation rules for industrial networks | |
| US20120063362A1 (en) | Method and apparatus for computing paths to destinations in networks having link constraints | |
| US20240428085A1 (en) | Isolation forest with ultra-low ram footprint for edge | |
| KR101739100B1 (ko) | 서비스 체이닝 가능한 오픈플로우 스위치 제어 방법 및 그 제어기 | |
| Zhang et al. | Towards automated network management: Learning the optimal protocol selection | |
| KR101707073B1 (ko) | Sdn 기반의 에러 탐색 네트워크 시스템 | |
| Mokrov et al. | Analytical model for software defined network considering memory node for routing rules | |
| Abbasi et al. | A genetic algorithm‐based flow update scheduler for software‐defined networks |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| 121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 17746907 Country of ref document: EP Kind code of ref document: A1 |
|
| WWE | Wipo information: entry into national phase |
Ref document number: 2017746907 Country of ref document: EP |
|
| NENP | Non-entry into the national phase |
Ref country code: DE |
|
| ENP | Entry into the national phase |
Ref document number: 2017746907 Country of ref document: EP Effective date: 20180731 |





















































































































