CN107995045B - Adaptive service function chain path selection method and system for network function virtualization - Google Patents

Adaptive service function chain path selection method and system for network function virtualization Download PDF

Info

Publication number
CN107995045B
CN107995045B CN201711377167.XA CN201711377167A CN107995045B CN 107995045 B CN107995045 B CN 107995045B CN 201711377167 A CN201711377167 A CN 201711377167A CN 107995045 B CN107995045 B CN 107995045B
Authority
CN
China
Prior art keywords
service function
service
path
unit
chain
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201711377167.XA
Other languages
Chinese (zh)
Other versions
CN107995045A (en
Inventor
马奕葳
陈俊良
周佳仪
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shanghai Maritime University
Original Assignee
Shanghai Maritime University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Shanghai Maritime University filed Critical Shanghai Maritime University
Priority to CN201711377167.XA priority Critical patent/CN107995045B/en
Publication of CN107995045A publication Critical patent/CN107995045A/en
Application granted granted Critical
Publication of CN107995045B publication Critical patent/CN107995045B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/50—Network service management, e.g. ensuring proper service fulfilment according to agreements
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/44—Arrangements for executing specific programs
    • G06F9/455—Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
    • G06F9/45533—Hypervisors; Virtual machine monitors
    • 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
    • H04L41/50—Network service management, e.g. ensuring proper service fulfilment according to agreements
    • H04L41/5041—Network service management, e.g. ensuring proper service fulfilment according to agreements characterised by the time relationship between creation and deployment of a service
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L63/00—Network architectures or network communication protocols for network security
    • H04L63/02—Network architectures or network communication protocols for network security for separating internal from external traffic, e.g. firewalls
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/50—Network services
    • H04L67/51—Discovery or management thereof, e.g. service location protocol [SLP] or web services
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/50—Network services
    • H04L67/56—Provisioning of proxy services
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/50—Network services
    • H04L67/60—Scheduling or organising the servicing of application requests, e.g. requests for application data transmissions using the analysis and optimisation of the required network resources
    • H04L67/63—Routing a service request depending on the request content or context

Abstract

The invention discloses an adaptive service function chain path selection method and system for network function virtualization, which comprises a service function chain controller; the service function chain controller is provided with a service function chain communication module and a service function path selection module; the service function chain communication module receives the request sent by the service requester and communicates a policy; the service function path selection module filters the service function instance and the service function path, calculates the benefit of the service function path, and selects the service function path with the best benefit. The invention provides a service function path selection mechanism, and the screening of the service function path can meet the required service chain type and reduce the load of each service function instance; the load of the service function instance can be dispersed, the excessive unbalance of the system environment load is avoided, and a service function path of a relatively high service level protocol can be provided; the invention increases the required number of service chains for receiving service and can reduce the packet loss rate.

Description

Adaptive service function chain path selection method and system for network function virtualization
Technical Field
The present invention relates to the field of network services, and in particular, to an adaptive service function chain path selection method and system for network function virtualization.
Background
With the rapid development of internet and cloud, network service providers have to add more dedicated hardware devices to provide specific network functions in a limited machine room space to meet the demand of providing more and more network services, and the connection and setup between the devices are very complicated. Therefore, in future networks, the traditional Network establishment method cannot meet the requirement of large growth, so the Network Function Virtualization (NFV) technology is a solution.
Network Function virtualization can virtualize network functions, place the network functions in a virtualization machine in a software manner to provide the network functions, and combine Service Function Chaining (SFC) technology to direct required Service functions one by one, thereby configuring the required network functions at will and reducing the complexity of network managers in building and managing network environments.
In an environment of network Function virtualization of a multi-Service Node (Service Node, Service Node SN), many Service functions are constructed, and in the face of a resource-limited environment and different network Function requirements, how to select a Service Function Path (Service Function Path, SFP) that should be combined by which Service Function Instances (Service Function Instances, SFIs), so as to avoid load imbalance in the system environment, and relatively provide a Service chain of a higher Service Level Agreement (SLA), which is an issue discussed in the present invention. Therefore, a service function chain integration architecture design and a service function path selection mechanism based on network function virtualization are needed to flexibly provide service chains so as to meet the service functions required by the service demanders.
With the continuous growth of internet technology, the type and mode of network services are changed, and a huge amount of data is accompanied; in order to solve the problems of storage, operation, and flow of these data, the current data processing method and network service method are oriented to virtualization and cloud-end, wherein the network function virtualization technology is developed rapidly, and the related technology and mechanism development of the service chain are also important for providing a lot of service functions.
The traditional network is built and configured in a static and fixed mode, and the network topology and the physical network are in a coupling mode, so that operators introduce new services and adjust network service chains, and can also actively connect and deploy the original network through a complicated configuration process to influence the original service supply, so that the traditional network service supply mode cannot meet a large number of network service requirements.
At present, network functions are virtualized in a dispute, a future network environment is likely to be built in a network function virtualization environment, the network functions can be realized on a virtual machine in a software mode, network function virtualization provides the capability of an uncoupled network, resources such as network topology and a physical network can be separated, network service configuration is more flexible and dynamic adjustment, and capital investment and maintenance cost of the overall maintenance of operators can be reduced.
In the data center, various service server clusters are constructed to provide different types of service requirements for service objects, and these different terminal service servers have different service chain requirements for service objects. As shown in fig. 1, for servers of different service types, the server has different service functions to be established between clients, for example, a proxy, a firewall and a network address translation service function are required between a Web server and a client, and a firewall and a video optimizer service function are required between a video server and a client; for the same server, it is desired to establish different service functions with different clients, for example, the server is more trusted for the client cluster 1, so it is desired to only pass through the firewall service function, and the client cluster 2 is less trusted, so it is desired to pass through the IPS (Intrusion Prevention System) service function for Prevention.
For servers with different service types, the server has different service functions to be established between clients, for example, a proxy, a firewall and a network address conversion service function are required between a Web server and the client, and a firewall and a video optimizer service function are required between a video server and the client; for the same server, it is desired to establish different service functions with different clients, for example, the server is more trusted for the client cluster 1, so it is desired to only pass through the firewall service function, and is less trusted for the client cluster 2, so it is desired to pass through the IPS (Intrusion prevention system) service function for prevention. In the face of the environment with limited resources and different network function requirements, further research is needed to provide a network service chain meeting the requirements of the service demander by using these resources.
In future network environments, to meet more and more service demands, it is possible to implement service function virtualization on virtual machines in software manner in a network function virtualization environment, where the network service functions are implemented on different service nodes. In an environment constructed by network function virtualization, there are many service nodes, on which many service functions are constructed, and different service nodes may have the same type of service functions, because different service nodes have different resources allocated to service function instances when the service function instances are initially constructed, and have different maintenance policies and stabilities, etc., the service level agreements provided by the service function instances of the same type of different service nodes are different, and thus each service function instance is ordered to have its corresponding price. A high price service function instance provides a relatively high service level agreement for the same type of service function instance. However, under the limited resources, how to select and adjust the service function paths to be combined by which service function instances, the service chain type and the allowable price condition that satisfy the service requirement, and the service chain that is given a relatively higher service level protocol under the load balance of a certain system environment are achieved. Therefore, the present invention provides a service chain that meets the requirement of a service requester by flexibly providing the service function chain in a network function virtualization environment, and therefore, the present invention is directed to design a service function chain control layer architecture and research and development of a service function routing mechanism.
Disclosure of Invention
The invention aims to provide an adaptive service function chain path selection method and system for network function virtualization, which designs the architecture and the operation flow of a service function chain controller system aiming at a service function chain based on network function virtualization, comprises two parts of service function chain communication and service function path selection, and mainly aims at a service function path selection module to realize a high-benefit and high-elasticity service chain supply system. The service function path selection module carries out service function instance overloading filtration and service function path excess filtration according to the requirements of the service demander, calculates the system benefit values of all service function paths one by one, and selects the service function path with the highest benefit for the system environment as a result. The service function path selection module can reduce the load of each service function instance in the environment of limited resources, avoid the excessive imbalance of the environmental load of the whole system and improve the network benefit of the whole system.
To achieve the above object, the present invention provides an adaptive service function chain path selection system for network function virtualization, which includes a service function chain controller for controlling and managing a service function chain formed by a client and a server in communication connection; the service function chain controller is provided with:
the service function chain communication module receives the request sent by the service requester connected with the service function chain communication module and communicates and issues a policy;
and the service function path selection module is connected with the service function chain communication module and is used for filtering the service function instances and the service function paths according to the service requester, calculating the benefits of the service function paths and selecting the service function path with the best benefit to provide for the service requester.
Preferably, the system comprises a control layer, a network virtualization layer and an infrastructure layer;
the service function chain controller is positioned at the control layer;
the network virtualization layer is provided with a virtual resource module, a virtual machine management program module, an inlet node, an outlet node and a plurality of service nodes; the virtual resource module comprises a virtual computing unit, a virtual storage unit and a virtual network unit; the inlet node and the outlet node are both provided with classifiers, and each service node is provided with a service function traffic forwarding unit and a plurality of service function units;
the hardware resources of the infrastructure layer include a computing unit, a storage unit, and a network unit.
Preferably, the service function chain communication module is provided with a receiving agent unit and a link policy unit;
the receiving agent unit receives the requirement of the service requester connected with the receiving agent unit, sends the requirement of the service function chain to the service function path selection module, and informs the service requester after the service function chain is established;
the link policy unit issues a policy to the classifier and the service function traffic forwarding unit to update rules according to the service function path selected by the service function path selection module so as to form a service function chain; wherein, the request sent by the service requester comprises: the source and destination information of the service chain to be established, the required service function chain type and the tolerable price.
Preferably, the service function path selection module is provided with a scheduling unit, a planning unit and a monitoring unit;
the monitoring unit monitors and collects the required service function instance information and provides the required service function instance information to the scheduling unit and the planning unit;
the dispatching unit is respectively connected with the receiving agent unit and the monitoring unit, receives the service function chain requirement from the receiving agent unit and acquires the required service function instance information from the monitoring unit, dispatches the service function instance information according to the requirement of the service demander and filters the service function instance and the service function path;
the planning unit is respectively connected with the scheduling unit and the monitoring unit, receives the service function paths which are sent by the scheduling unit and reserved after filtering, calculates the benefit of each service function path reserved after filtering, selects the service function path with the highest benefit value as a result, and provides the service function path result to the monitoring unit.
Preferably, the planning unit is connected to the link policy unit, the planning unit notifies the link policy unit to issue policy update rules after obtaining the selected service function path result, the link policy unit notifies the receiving agent unit to complete the establishment of the service function chain after having assigned the rules, and the receiving agent unit notifies the service requester that the service function chain is established.
Preferably, the service function instance information includes: CPU usage, memory usage, network throughput, maximum network throughput, subscription price, and ID of the corresponding service function instance.
The invention also provides an adaptive service function chain path selection method of the adaptive service function chain path selection system based on the network function virtualization, which comprises the following steps:
step S1, the service function chain controller receives the requirement of the service requester;
step S2, a service function path selection module in the service function chain controller filters all service function paths, eliminates the service function paths containing the combination of the heavy-load service function instances and the service function paths exceeding the limit price, and obtains the service function paths reserved after filtering;
step S3, the service function path selection module calculates the respective benefit of the reserved service function path, and takes the service function path with the highest benefit value as the result;
step S4, the service function chain communication module in the service function chain controller executes a policy update rule according to the selected service function path result, and informs the service requester that the service requester has completed establishing the required service function chain.
Preferably, the service function path selection module is provided with a scheduling unit, a monitoring unit and a planning unit;
the step S2 includes:
s21: the dispatching unit requests the monitoring unit for the information of all service function instances of the service function types of the service function chain types meeting the requirements of the service provider;
s22: the monitoring unit returns the required service function instance information to the scheduling unit;
s23: the scheduling unit calculates the resource utilization rate of each service function instance, filters out a service function path containing any reloading service function instance, lists the rest service function paths, calculates the price of each service function path in the rest service function paths, and eliminates the service function paths exceeding the limit price;
s24: when the number of the service function paths reserved after filtering is less than 0, the service function chain controller refuses to supply the service requester with the service function chain requirement; or, when the number of the service function paths reserved after filtering is greater than 0, the scheduling unit transmits the reserved service function path information to the planning unit.
Preferably, the step S23 includes:
a. filtering out the service function path containing the service function instance in the overload state by calculating the resource utilization rate of each service function instance when the resource utilization rate of the service function instance
Figure BDA0001514896630000061
Less than a threshold R for tolerable resource usageTTIf the current time is within the acceptable range, otherwise, the current time is in a heavy load state; calculating the resource utilization rate of the service function instance according to formula (1), formula (2) and formula (3):
Figure BDA0001514896630000062
Figure BDA0001514896630000063
WC+WM+WT＝1,WC,WM,WT≥0 (3)
in the formula (I), the compound is shown in the specification,
Figure BDA0001514896630000064
indicating service function as FlCPU usage of the ith service function instance of (2),
Figure BDA0001514896630000065
indicating service function as FlThe memory usage of the ith service function instance,
Figure BDA0001514896630000066
representing service functionsIs FlUsing network throughput percentage, W, of the ith service function instance of (1)CIs the weighted value of the CPU utilization rate to the resource utilization rate, WMIs the weighted value of the memory usage rate to the resource usage rate, WTA weighted value of the network throughput utilization rate in the resource utilization rate;
Figure BDA0001514896630000067
as a service function FlThe network throughput of the i-th SFI,
Figure BDA0001514896630000068
as a service function FlMaximum network throughput of the ith service function instance of (1);
b. filtering the service function paths exceeding the limit price in the rest service function paths when the total price of one service function path
Figure BDA0001514896630000069
Allowable price P greater than service requester's requirementLIf yes, the service function path is removed;
calculating the total price of the service function path according to formula (4):
Figure BDA00015148966300000610
in the formula (I), the compound is shown in the specification,
Figure BDA00015148966300000611
is composed of
Figure BDA00015148966300000612
The total price of (a) is,
Figure BDA00015148966300000613
is SFPDMiddle jth service function path, SFPDA service function path set reserved after the service function path containing the overloaded service function instance is filtered; n is a radical ofsFor the length of the required service function chain, PLThe allowable price required by the service requester.
Preferably, the step S3 includes:
calculating the benefit of the service function path according to formula (5):
Figure BDA0001514896630000071
Figure BDA0001514896630000072
WR+WP＝1,WR,WP≥0 (7)
in the formula (I), the compound is shown in the specification,
Figure BDA0001514896630000073
represents
Figure BDA0001514896630000074
The benefit of (2);
Figure BDA0001514896630000075
represents
Figure BDA0001514896630000076
The total resource utilization rate is the sum of the service function instance resource utilization rates of the service function path; ,
Figure BDA0001514896630000077
represents
Figure BDA0001514896630000078
Middle FlResource usage of; wRAnd WPWeight values respectively representing the resource rate and the benefit occupied by the price;
Figure BDA0001514896630000079
the usable resource rate is expressed, and the higher the usable resource rate, the lighter the load of the current service function path is;
Figure BDA00015148966300000710
the price and the usable resource utilization rate are in the same measurement proportion of 0-100%, and the closer to 100% represents that the price of the current service function path is closer to the allowable price required by the service requester.
Compared with the prior art, the invention has the beneficial effects that: (1) the invention relates to a service function chain system based on a network function virtualization environment, which combines network function virtualization and a service function chain concept architecture. (2) The invention designs a service function chain controller, and provides an internal module structure and an operation flow, thereby achieving the purpose of automatically supplying the service function chain. (3) The invention provides a service function path selection mechanism: the screening service function path can meet the service chain type required by the service demander, and can reduce the load of each service function instance; the invention can also disperse the load of the service function instance, and avoid the condition that the environmental load of the whole system is too unbalanced, and the system can provide a service function path of a relatively high service level protocol; the invention increases the required number of service chains for receiving service and can reduce the packet loss rate.
Drawings
FIG. 1 is a schematic diagram of different service chain requirements;
FIG. 2 is a schematic diagram of a service function chain path selection system based on network function virtualization according to the present invention;
FIG. 3 is a schematic diagram of a hierarchy of a service function chain routing system according to the present invention;
FIG. 4 is a schematic diagram of a network function virtualization environment of the present invention;
FIG. 5 is a schematic diagram illustrating the overall operation of the service function chain routing system according to the present invention;
FIG. 6 is a schematic diagram of a method for operating a service function path selection module according to the present invention;
FIG. 7 is a schematic diagram of a resource usage threshold in accordance with the present invention;
FIG. 8 is a flow diagram of a service function path selection module filtering a reload service function instance in accordance with the present invention;
FIG. 9 is a flow diagram of the present invention for filtering service function paths that exceed a limit price;
FIG. 10 is a flow chart of the present invention for obtaining the most efficient service function path.
Detailed Description
The present invention provides an adaptive service function chain path selection method and system for network function virtualization, and the present invention is further described below with reference to the accompanying drawings and the detailed description for the purpose of making the present invention more obvious and easy to understand.
The invention provides a service function chain controller system based on a service function chain of network function virtualization, which comprises two parts of service function chain communication and service function path selection, and mainly aims at a service function path selection method to realize a high-benefit and high-elasticity service chain supply system.
As shown in fig. 2, the service function chain path selection system based on network function virtualization of the present invention mainly includes a service function chain controller, a server and a client. The service function chain is formed by establishing the service function between the server and the client, and the service function path between the server and the client has a plurality of different choices, such as a service function path 1, a service function path 2, a service function path k, and the like. The service function chain controller is used for managing and controlling the service function chain. The service function chain comprises a plurality of service nodes, and each service node is provided with a firewall or an intrusion prevention system.
The action flow from the service requester requiring the service to the service requester arranging the service function path is specifically as follows: the network manager will establish the connection between the server and the client according to the service requirement between the server and a certain client, so as to request the service function chain (the service function chain is the firewall and the intrusion prevention system) to the service function chain controller (i.e. the service function chain system manager and controller), and provide the service requirement message to the service function chain controller, and the service function chain controller will arrange the service function path meeting the requirement and having the highest system benefit to the server through the service function path selection mechanism according to the message.
Wherein, the service requirement of the service requester comprises: source (client information, e.g., the client's IP), destination (server information, e.g., the server's IP), service function chain (firewall and intrusion prevention system), type of network function required and allowable price plan, etc.
As shown in fig. 2 and fig. 3 in combination, the service function chain routing system based on network function virtualization of the present invention mainly includes a control layer, a network virtualization layer and an infrastructure layer.
The control layer realizes service function chain communication and service function path selection through the service function chain controller. The service function chain controller is provided with a service function chain communication module and a service function path selection module.
The service function chain communication module is provided with a receiving agent unit and a link policy unit and is used for receiving the request, communicating and issuing the policy. The receiving agent unit interactively communicates with a request server (namely a requester), and acquires the information of the request service function chain sent by the request server.
The service function path selection module can perform heavy load filtering of the service function instance and excess filtering of the service function path according to the requirement of the service demander, calculate the system benefit value of each service function path one by one, and select the service function path with the highest benefit for the system environment as a result. Namely, the service function path selection module arranges a service function path with the best system benefit at present from different service function instances to provide for the service demander.
The service function path selection module is provided with a scheduling unit, a planning unit and a monitoring unit, so that a service function path which meets the conditions and has the best system benefit is selected according to the requirements of requesters.
The dispatching unit is connected with the receiving agent unit and the monitoring unit respectively, and obtains required information from the receiving agent unit and the monitoring unit (for example, the dispatching unit receives a service function chain requirement from the receiving agent unit and the dispatching unit needs to obtain service function instance information of a service function type in the service function chain type meeting the requirement of a service requester in the environment, such as information of CPU utilization rate, memory utilization rate, network throughput, maximum network throughput, a set price and an ID of a corresponding service function instance from the monitoring unit), dispatches the information of the service function instance according to the requirement of the service requester, and calculates and filters out a service function path containing a heavy load service function instance combination and a service function path exceeding a limit price according to the current service function instance information and conditions.
The planning unit is respectively connected with the scheduling unit and the monitoring unit, receives the service function paths which are sent by the scheduling unit and reserved after filtering, calculates the benefit value of each service function path reserved after filtering, selects the service function path with the highest benefit value as the task result, and provides the service function path result to the monitoring unit.
The planning unit is also connected with the link policy unit, the planning unit informs the link policy unit to issue policy updating rules after obtaining the result of the selected service function path, the link policy unit informs the receiving agent unit that the task is completed after the link policy unit finishes assigning the rules, and the receiving agent unit informs the service requester that the service chain is established.
A monitoring unit: monitoring and collecting the required service function instance information, and providing the required service function instance information to the scheduling unit and the planning unit.
Meanwhile, the receiving agent unit is responsible for communicating with the service requester and the function module of the service function chain, receiving the request of the service requester, forwarding the request of the service function chain to the scheduling unit, and notifying the requester after the service function chain is established.
The link policy unit is used for issuing a policy to the classifier and the service function traffic forwarding unit for updating rules according to the selected service function path so as to realize a service function chain.
The network virtualization layer is provided with a virtual resource module, a virtual machine manager module, an entry node, an exit node and a plurality of service nodes (for example, service node 1, service node 2, service node n, etc.).
The ingress node receives data messages from the source, and the egress node transmits data messages to the destination. The service node is a collection node for storing service functions. The inlet node and the outlet node are provided with classifiers. Each service node is provided with a service function traffic forwarding unit and a plurality of service function units.
The virtual resource module mainly comprises a virtual computing unit, a virtual storage unit and a virtual network unit. The virtual resource module is mainly used for integrating hardware resources of an infrastructure layer to form complete virtual resources, and the complete virtual resources can be used as upper-layer service nodes. The virtual machine management program module is mainly responsible for managing the virtual machine, and the virtual machine is stored in the virtual resource module environment.
The hardware resources of the infrastructure layer mainly comprise a computing unit, a storage unit and a network unit. The computing unit is used by the computer environment to compute and process information. The storage unit is used for storing information for the computer environment. The network element serves to communicate information for the computer environment.
As shown in fig. 3 and 4, the service function chain controller of the present invention plans a service function chain path according to a required service function chain, and issues a policy to the service function chain element according to the planned service function chain path, and updates the classification and forwarding policy rules, etc., so that when a conforming information flow enters, a packet is transmitted according to the service function chain path.
The network function virtualization environment comprises a plurality of service nodes, one service node is equivalent to one or more servers internally provided with a plurality of virtual machines, and one service function unit is loaded on one virtual machine. All the service function instances are connected to the service function traffic forwarding units, the service function traffic forwarding units in different service nodes are connected with each other, and classifiers are built at the ingress node and the egress node to form a service function chain path selection system bottom layer based on network function virtualization. Under different service offerings, the service function chain will be different, and therefore the service function path will be separated.
The service function chain path selection system of the invention knows the corresponding positions of all service function chain elements, and the bottom layer connection position can not be changed. If the resources in the network function virtualization environment have no additional added devices or unallocated resources, the overall resources are limited, and the resources of the individual service function instance are allocated at the beginning when the service function unit is built and loaded. All packets entering the system environment will pass through the classifier, which will add header or label to them, the service function traffic forwarding unit will forward the packets according to the added header information or label, and the next connected service function unit or down-transmit to other service function traffic forwarding units according to the forwarding policy rules, and finally the packets will be passed out of the system environment by removing the added header or label through the classifier.
As shown in fig. 4, the service function chain controller plans a service function path according to the required service function chain, and issues a policy to the service function chain element according to the planned service function path, and updates the classification and forwarding policy rules, so that when the conforming information flow enters, the packet is transmitted according to the service function path. The present invention focuses on service function path selection by the service function chain controller to select the eligible and currently most appropriate service function path.
The request information obtained by the service function chain controller from the service requester is shown in table 1, and the service requester provides the service function chain controller with the source and destination information that the service requester intends to establish a service chain, and the service function chain type and acceptable price information. The service function chain controller will select the better service function path according to the service function chain type based on these information, and provide the service chain by adding or updating the rule.
TABLE 1 request information Table obtained from service requester
Figure BDA0001514896630000111
The overall operation flow of the system of the service function chain controller of the invention is as follows:
after the service function chain controller receives the service function chain request message, a service function path which meets the conditions at present and is better is determined through the service function path selection module. The specific process comprises the following steps:
in step S1, the service function chain controller receives a request message of the service requester, including the source and destination information of the service chain to be established, the required service function chain type and the tolerable price.
And step S2, the service function path selection module filters all service function paths and eliminates the service function paths containing the heavy-load service function instance combination and the service function paths exceeding the limit price.
Step S3, the service function path selection module calculates the system benefit values of the respective reserved service function paths, and obtains the service function path with the highest benefit value as the result.
Step S4, the service function chain communication module executes the policy updating rule according to the selected service function path, and informs the service requester that the service function chain for the request has been established.
FIG. 5 is a schematic diagram of the overall operation of the service function chain path selection system according to the present invention, in which the service function controller receives and schedules service requests from the service function chain receiving agent unit, and sequentially sends tasks to the service function path selection module to screen service function paths through the service function path selection module and select a better service function path. In the operation of the service function path selection module, the scheduling unit needs to acquire service function instance information (such as CPU utilization, memory utilization, network throughput, maximum network throughput, a predetermined price, and ID of a corresponding service function instance) of a service function type in a service function chain type meeting the requirements of a service requester in an environment from the monitoring unit, the monitoring unit returns the collected information to the scheduling unit, and the scheduling unit calculates and filters a service function path including a combination of reloading service function instances and a service function path exceeding a limit price. The planning unit calculates the benefit value of each service function path reserved after filtering, and selects the service function path with the highest benefit value as the task result; after the result of the selected service function path is obtained, the planning unit informs the link policy unit to issue a policy updating rule, after the link policy module finishes assigning the rule, the receiving agent unit is informed that the task is finished, and the receiving agent unit informs the service requester that the service chain is established.
When a service function chain request is received, the service function chain controller decides a service function path through a service function path selection module; as shown in fig. 6, the steps of the operation method of the service function path selection module are as follows:
step T1, service function instance resource information in the scheduling environment: the dispatch unit requests information of all service function instances of the service function type that matches the service function chain type requested by the server from the monitor unit.
In step T2, the monitoring unit returns the requested service function instance information to the dispatch unit.
And step T3, the scheduling unit calculates the resource utilization rate of each service function instance, filters and eliminates the service function paths containing any reloaded service function instance, lists the rest service function paths, calculates the price of each service function path in the rest service function paths, and eliminates the service function paths exceeding the limit price. Then judging whether the number of the service function paths reserved at present is more than 0: if not, it means that all service function paths are deleted, i.e. no service function path is selected, which means that the current environment cannot bear the requirement or the requirement may cause the system environment to be under a network condition, so the system refuses to provide the service function chain requirement. If yes, the dispatching unit transmits the reserved path information of each service function to the planning unit.
And step T4, the planning unit calculates the received benefit value of each service function path currently reserved, and selects the service function path with the highest benefit value as the result.
Specifically, the monitoring unit monitoring information process in step T1 is: initially, service function instance resource information in the environment is scheduled to obtain the service function type of the service function chain meeting the requirement of the service requester from the monitoring unitService function instance information. The service function chain that defines the service requestor requirements here is labeled:
Figure BDA0001514896630000132
Flrepresenting the service function of the l-th order in the service function chain, NsRepresenting the number of service functions in the service function chain. Assuming that the service function chain requested by the service requester is s: { SF 1, SF 2}, it indicates F1Is SF 1, F2Is SF 2 and NsThe value is 2, and the obtained monitoring information is the service function instance information of all service function types SF 1 and SF 2 in the environment as shown in table 2 and table 3.
TABLE 2 service function instance information in the Environment
Figure BDA0001514896630000131
TABLE 3 service function instance information in the Environment
Figure BDA0001514896630000141
Specifically, the service function path process of filtering the service function instance containing the reload status in step T3 is as follows: after acquiring the information of the corresponding service function instances required in the environment, calculating the resource utilization rate of each service function instance and filtering the service function instances in the overload state, thereby avoiding the situations of insufficient processing or congestion and the like caused by the fact that the selected service function path contains the service function instances in the overload state to provide services. Figure 7 is a diagram of resource usage thresholds,
Figure BDA0001514896630000142
representative service function type is FlResource utilization, R, of the ith service function instance ofTTIs an allowable resource usage threshold if the resource usage of the service function instance is less than RTTIf the current is in the acceptable range, otherwise, the current is in the heavy load state; including any service function instance in a overloaded stateNone of the combined service function paths are included in the selection, and FIG. 8 is a flow chart illustrating the filtering of the reload service function instance by the service function path selection module. Equation (1) may be utilized to calculate the resource usage scenario for each service function instance,
Figure BDA0001514896630000143
indicating service function as FiThe CPU usage rate of the ith service function instance;
Figure BDA0001514896630000144
indicating service function as FlThe memory usage rate of the ith service function instance;
Figure BDA0001514896630000145
indicating service function as FlThe network throughput percentage of the ith service function instance of (1); wCThe CPU utilization rate is the weighted value of the resource utilization rate occupied by the CPU utilization rate; wMThe weight value of the memory utilization rate accounting for the resource utilization rate; wTIs the weighted value of the network throughput utilization rate and the resource utilization rate. The percentage of network throughput used may be calculated using equation (2)
Figure BDA0001514896630000146
As a service function FlNetwork throughput of the ith SFI;
Figure BDA0001514896630000147
as a service function FlMaximum network throughput of the ith service function instance of (1).
Figure BDA0001514896630000148
Figure BDA0001514896630000149
WC+WM+WT＝1,WC,WM,WT≥0 (3)
Wherein, WC、WMAnd WTThe weight values of the CPU utilization, the memory utilization, and the percentage of the network throughput used to the resource utilization are respectively expressed, as shown in formula (3), the total limit value is 1, and each weight value needs to be greater than or equal to 0. System personnel can adjust each weight value according to the requirements and importance of the system personnel on the network environment.
Specifically, the service function path process of filtering exceeding the limit price in step T3 is: after the service function paths containing the overloaded service function instances are filtered, the rest service function paths are listed, the total price of each service function path is calculated, and the service function paths exceeding the limit price are removed, so that the provided service function paths are prevented from not meeting the allowable price of a service requester. Fig. 9 is a flow chart showing the service function path selection module filtering the service function path exceeding the limit price. Here, SFP is labeledDTo filter the set of service function paths that remain after the service function path containing the service function instance that is overloaded,
Figure BDA0001514896630000151
is SFPDThe jth service function path;
Figure BDA0001514896630000152
is composed of
Figure BDA0001514896630000153
The total price of the service function path can be calculated using equation (4),
Figure BDA0001514896630000154
is composed of
Figure BDA0001514896630000155
Middle FlPrice of, NsThe length of the required service function chain (i.e., the number of service function types within the required service function chain); pLTo limit the price (i.e., the allowable price requested by the service requester).
Figure BDA0001514896630000156
If greater than the limit price PLThen the service function path is eliminated.
Figure BDA0001514896630000157
Specifically, the process of obtaining the service function path with the highest benefit value in step T4 is as follows: after removing the service function paths exceeding the limit price, the benefit of each path in all the remaining service function paths is calculated first to obtain the maximum benefit value, and the service function path with the maximum benefit value is selected as the selection result, so as to obtain a service function path with the best benefit for the system, that is, the service function path with the highest benefit is selected as the result, which can avoid the system from being too unbalanced, and can provide a relatively high service level protocol, for example, the flow chart of the service function path with the highest benefit is obtained by the service function path selection module shown in fig. 10.
In the system environment, the lower the resource usage rate and the higher the price of the service function path are selected and allocated, the higher the benefit is, so the benefit is calculated in consideration of the resource usage rate and the price of the service function path.
The benefit of the service function path can be calculated using equation (5),
Figure BDA0001514896630000158
represents
Figure BDA0001514896630000159
The benefit of (2);
Figure BDA00015148966300001510
represents
Figure BDA0001514896630000161
The total resource utilization of this service function path, i.e. the sum of the service function instance resource utilization of this service function path, can be obtained from equation (6),
Figure BDA0001514896630000162
represents
Figure BDA0001514896630000163
Middle FlResource usage of, NsThe length of the required service function chain (i.e., the number of service function types within the required service function chain);
Figure BDA0001514896630000164
is composed of
Figure BDA0001514896630000165
The total price of (c); pLTo limit the price (i.e., the allowable price requested by the service requester).
Wherein the content of the first and second substances,
Figure BDA0001514896630000166
the higher the resource rate is, the lighter the load of the current service function path is;
Figure BDA0001514896630000167
the price and the usable resource utilization rate are in a same measurement proportion of 0-100%, and the closer to 100% represents that the price of the service function path is closer to the limit price, the better the service level protocol can be provided; wRAnd WPThe weighted values of the benefit of the resource rate and the price, respectively, are shown in equation (7), the total limit value is 1, and each weighted value needs to be greater than or equal to 0. System personnel can adjust each weight value according to the requirements and importance of the system personnel on the network environment.
Figure BDA0001514896630000168
Figure BDA0001514896630000169
WR+WP＝1,WR,WP≥0 (7)
To sum up, the method for selecting a service function path of the present invention mainly calls the corresponding service function instance information required in the environment according to the service function chain type required by the service requester, filters out the service function path containing any combination of service function instances in the overloaded state, calculates the prices of the remaining service function paths, eliminates the service function paths that do not meet the requirements of the service requester, and then calculates the individual benefit values of the remaining service function paths, thereby selecting the service function path with the best benefit for the system environment and avoiding the load from being unbalanced. If the system of the present invention is not introduced, the establishment mode of the network function service chain is not selected in a better path planning manner, and whether the network function service chain meets the requirements of the service demanders is not considered, which will cause that the virtual machine providing the network function service is easy to enter a heavy load state, thereby causing poor network benefits of the system and possibly failing to meet the requirements of the service demanders. Therefore, the invention configures the network function through the network virtualization mode, under the limited network resource, according to the requirement of the service demander, analyzes the network environment condition, and plans a service function path with the maximum system benefit to provide for the service demander, thereby avoiding the load imbalance, providing a higher service level protocol, and reducing the cost requirement on the network function configuration and the complexity on the configuration compared with the traditional network. For a large amount of network function requirements in the future, the flexible service chain architecture and the service function routing mechanism can achieve the purpose of strengthening the overall network benefit.
While the present invention has been described in detail with reference to the preferred embodiments, it should be understood that the above description should not be taken as limiting the invention. Various modifications and alterations to this invention will become apparent to those skilled in the art upon reading the foregoing description. Accordingly, the scope of the invention should be determined from the following claims.

Claims (7)

1. An adaptive service function chain path selection system for network function virtualization, comprising a service function chain controller for controlling and managing a service function chain formed by a client and a server communicatively connected; the service function chain controller is provided with:
the service function chain communication module receives the request sent by the service requester connected with the service function chain communication module and communicates and issues a policy;
the service function path selection module is connected with the service function chain communication module and is used for filtering the service function instances and the service function paths according to the requirements of the service demander, calculating the benefits of the service function paths and selecting the service function path with the best benefit to provide for the service demander;
the adaptive service function chain path selection system comprises a control layer, a network virtualization layer and an infrastructure layer;
the service function chain controller is positioned at the control layer;
the network virtualization layer is provided with a virtual resource module, a virtual machine management program module, an inlet node, an outlet node and a plurality of service nodes; the virtual resource module comprises a virtual computing unit, a virtual storage unit and a virtual network unit; the inlet node and the outlet node are both provided with classifiers, and each service node is provided with a service function traffic forwarding unit and a plurality of service function units;
the hardware resources of the infrastructure layer comprise a computing unit, a storage unit and a network unit; the service function chain communication module is provided with a receiving agent unit and a link policy unit;
the receiving agent unit receives the requirement of the service requester connected with the receiving agent unit, sends the requirement of the service function chain to the service function path selection module, and informs the service requester after the service function chain is established;
the link policy unit issues a policy to the classifier and the service function traffic forwarding unit to update rules according to the service function path selected by the service function path selection module so as to form a service function chain;
wherein, the request sent by the service requester comprises: the source and destination information of the service chain to be established, the required service function chain type and the tolerable price;
the service function path selection module is provided with a scheduling unit, a planning unit and a monitoring unit; the monitoring unit monitors and collects the required service function instance information and provides the required service function instance information to the scheduling unit and the planning unit;
the dispatching unit is respectively connected with the receiving agent unit and the monitoring unit, receives the service function chain requirement from the receiving agent unit and acquires the required service function instance information from the monitoring unit, dispatches the service function instance information according to the requirement of the service demander and filters the service function instance and the service function path;
the planning unit is respectively connected with the scheduling unit and the monitoring unit, receives the service function paths which are sent by the scheduling unit and reserved after filtering, calculates the benefit of each service function path reserved after filtering, selects the service function path with the highest benefit value as a result, and provides the service function path result to the monitoring unit.
2. The adaptive service function chain path selection system of claim 1, wherein the planning unit is coupled to the link policy unit, the planning unit notifies the link policy unit to issue policy update rules after obtaining the selected service function path result, the link policy unit notifies the receiving agent unit to complete the service function chain establishment after assigning the rules, and the receiving agent unit notifies the service requester that the service function chain establishment is complete.
3. The adaptive service function chain path selection system of claim 1, wherein the service function instance information comprises: CPU usage, memory usage, network throughput, maximum network throughput, subscription price, and ID of the corresponding service function instance.
4. An adaptive service function chain path selection method based on the adaptive service function chain path selection system for network function virtualization according to any one of claims 1 to 3, the method comprising the steps of:
step S1, the service function chain controller receives the requirement of the service requester;
step S2, a service function path selection module in the service function chain controller filters all service function paths, eliminates the service function paths containing the combination of the heavy-load service function instances and the service function paths exceeding the limit price, and obtains the service function paths reserved after filtering;
step S3, the service function path selection module calculates the respective benefit of the reserved service function path, and takes the service function path with the highest benefit value as the result;
step S4, the service function chain communication module in the service function chain controller executes a policy update rule according to the selected service function path result, and informs the service requester that the service requester has completed establishing the required service function chain.
5. The adaptive service function chain path selection method of claim 4, wherein the service function path selection module is provided with a scheduling unit, a monitoring unit, and a planning unit;
the step S2 includes:
s21: the dispatching unit requests the monitoring unit for the information of all service function instances of the service function types which accord with the service function chain type of the service requester;
s22: the monitoring unit returns the required service function instance information to the scheduling unit;
s23: the scheduling unit calculates the resource utilization rate of each service function instance, filters out a service function path containing any reloading service function instance, lists the rest service function paths, calculates the price of each service function path in the rest service function paths, and eliminates the service function paths exceeding the limit price;
s24: when the number of the service function paths reserved after filtering is less than 0, the service function chain controller refuses to supply the service requester with the service function chain requirement; or, when the number of the service function paths reserved after filtering is greater than 0, the scheduling unit transmits the reserved service function path information to the planning unit.
6. The adaptive service function chain path selection method of claim 5,
the step S23 includes:
a. filtering out the service function path containing the service function instance in the overload state by calculating the resource utilization rate of each service function instance when the resource utilization rate of the service function instance
Figure FDA0002633695360000031
Less than a threshold R for tolerable resource usageTTIf the current time is within the acceptable range, otherwise, the current time is in a heavy load state;
calculating the resource utilization rate of the service function instance according to formula (1), formula (2) and formula (3):
Figure FDA0002633695360000032
Figure FDA0002633695360000033
WC+WM+WT＝1，WC，WM，WT≥0 (3)
in the formula (I), the compound is shown in the specification,
Figure FDA0002633695360000034
indicating service function as FlCPU usage of the ith service function instance of (2),
Figure FDA0002633695360000035
indicating service function as FlThe memory of the ith service function instance of (1)The rate of utilization is higher than the rate of utilization,
Figure FDA0002633695360000036
indicating service function as FlUsing network throughput percentage, W, of the ith service function instance of (1)CIs the weighted value of the CPU utilization rate to the resource utilization rate, WMIs the weighted value of the memory usage rate to the resource usage rate, WTA weighted value of the network throughput utilization rate in the resource utilization rate;
Figure FDA0002633695360000037
as a service function FlThe network throughput of the i-th SFI,
Figure FDA0002633695360000041
as a service function FlMaximum network throughput of the ith service function instance of (1);
b. filtering the service function paths exceeding the limit price in the rest service function paths when the total price of one service function path
Figure FDA0002633695360000042
Allowable price P greater than service requester's requirementLIf yes, the service function path is removed;
calculating the total price of the service function path according to formula (4):
Figure FDA0002633695360000043
in the formula (I), the compound is shown in the specification,
Figure FDA0002633695360000044
is composed of
Figure FDA0002633695360000045
The total price of (a) is,
Figure FDA0002633695360000046
i.e. the jth service function path in SFPD, SFPDA service function path set reserved after the service function path containing the overloaded service function instance is filtered; n is a radical ofsFor the length of the required service function chain, PLThe allowable price required by the service requester.
7. The adaptive service function chain path selection method of claim 6,
the step S3 includes:
calculating the benefit of the service function path according to formula (5):
Figure FDA0002633695360000047
Figure FDA0002633695360000048
WR+WP＝1，WR，WP≥0 (7)
in the formula (I), the compound is shown in the specification,
Figure FDA0002633695360000049
represents
Figure FDA00026336953600000410
The benefit of (2);
Figure FDA00026336953600000411
represents
Figure FDA00026336953600000412
The total resource utilization rate is the sum of the service function instance resource utilization rates of the service function path;
Figure FDA00026336953600000413
represents
Figure FDA00026336953600000414
Middle FlResource usage of; wRAnd WPWeight values respectively representing the resource rate and the benefit occupied by the price;
Figure FDA00026336953600000415
the usable resource rate is expressed, and the higher the usable resource rate, the lighter the load of the current service function path is;
Figure FDA00026336953600000416
the price and the usable resource utilization rate are in the same measurement proportion of 0-100%, and the closer to 100% represents that the price of the current service function path is closer to the allowable price required by the service requester.
CN201711377167.XA 2017-12-19 2017-12-19 Adaptive service function chain path selection method and system for network function virtualization Active CN107995045B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201711377167.XA CN107995045B (en) 2017-12-19 2017-12-19 Adaptive service function chain path selection method and system for network function virtualization

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201711377167.XA CN107995045B (en) 2017-12-19 2017-12-19 Adaptive service function chain path selection method and system for network function virtualization

Publications (2)

Publication Number Publication Date
CN107995045A CN107995045A (en) 2018-05-04
CN107995045B true CN107995045B (en) 2020-10-13

Family

ID=62038872

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201711377167.XA Active CN107995045B (en) 2017-12-19 2017-12-19 Adaptive service function chain path selection method and system for network function virtualization

Country Status (1)

Country Link
CN (1) CN107995045B (en)

Families Citing this family (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110830538B (en) * 2018-08-13 2022-06-14 华为技术有限公司 Message transmission method, device and storage medium
US10805221B2 (en) 2018-11-06 2020-10-13 Nanning Fugui Precision Industrial Co., Ltd. Service function chain (SFC) path selection method and system
US10873528B2 (en) * 2018-11-24 2020-12-22 Nanning Fugui Precision Industrial Co., Ltd. Packet forwarding method and device utilizing the same
CN109547439B (en) * 2018-11-26 2021-04-27 奇安信科技集团股份有限公司 Processing method and device for service node access network
CN109787813B (en) * 2018-12-21 2022-02-22 奇安信科技集团股份有限公司 SDN-based service chain configuration method, device, system and medium
US10944647B2 (en) * 2019-01-24 2021-03-09 Vmware, Inc. Dynamic inter-cloud placement of virtual network functions for a slice
US11012288B2 (en) 2019-05-14 2021-05-18 Vmware, Inc. Congestion avoidance in a slice-based network
US11588733B2 (en) 2019-05-14 2023-02-21 Vmware, Inc. Slice-based routing
US10892994B2 (en) 2019-05-14 2021-01-12 Vmware, Inc. Quality of service in virtual service networks

Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107483286A (en) * 2017-08-14 2017-12-15 电子科技大学 Merge the method with deployment services functional chain under a kind of environment based on cloud and mist

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9462084B2 (en) * 2014-12-23 2016-10-04 Intel Corporation Parallel processing of service functions in service function chains

Patent Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107483286A (en) * 2017-08-14 2017-12-15 电子科技大学 Merge the method with deployment services functional chain under a kind of environment based on cloud and mist

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
数据中心网络中服务功能链的部署与路由;杨柯;《中国优秀硕士学位论文全文数据库（信息科技辑）》;20171115(第11期);第12-14，31-37页 *

Also Published As

Publication number Publication date
CN107995045A (en) 2018-05-04

Similar Documents

Publication Publication Date Title
CN107995045B (en) Adaptive service function chain path selection method and system for network function virtualization
CN103207814B (en) Managing and task scheduling system and dispatching method across cluster resource of a kind of decentration
EP1851921B1 (en) Admission control and routing in a telecommunications network based on the consideration of all physical and logical links
Ghaznavi et al. Service function chaining simplified
US6459682B1 (en) Architecture for supporting service level agreements in an IP network
US20030236887A1 (en) Cluster bandwidth management algorithms
Bu et al. Enabling adaptive routing service customization via the integration of SDN and NFV
US20020152305A1 (en) Systems and methods for resource utilization analysis in information management environments
US6857025B1 (en) Highly scalable system and method of regulating internet traffic to server farm to support (min,max) bandwidth usage-based service level agreements
US20040153545A1 (en) Software, systems and methods for managing a distributed network
CN106533806B (en) Method for providing cross-layer QoS based on application awareness in multi-tenant SDN network
CN104798356A (en) Method and apparatus for controlling utilization in a horizontally scaled software application
CN109787801A (en) A kind of network service management methods, devices and systems
Sharma et al. VNF availability and SFC sizing model for service provider networks
JP2000312226A (en) Method for warranting communication quality
Ma et al. Adaptive service function selection for Network Function Virtualization networking
CN107347089A (en) A kind of resource allocation methods of carrier-class cloud computing system
Dong et al. A novel distribution service policy for crowdsourced live streaming in cloud platform
Abdelaal et al. Network-aware resource management strategy in cloud computing environments
Lee et al. Enhancing infiniband with openflow-style sdn capability
Li et al. CoMan: Managing bandwidth across computing frameworks in multiplexed datacenters
JP2005182702A (en) Access control system in ip network
US6928053B1 (en) Dynamic resource control in telecommunications networks
Bunyakitanon et al. HELICON: Orchestrating low-latent & load-balanced Virtual Network Functions
Romanov et al. Principles of building modular control plane in software-defined network

Legal Events

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