JP2004274368A - Quality guarantee controller and load distributing device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To reduce the burden of quality guarantee control by a service provider. <P>SOLUTION: When a quality guarantee server receives a use request for a network service including a quality guarantee request, the quality guarantee server detects the use frequency level of a service per hour and sets a QoS path which can house traffic flow to the use request and collect traffic flow to a use request that is generated thereafter as provisioning of quality guarantee control corresponding to the use frequency level. <P>COPYRIGHT: (C)2004,JPO&NCIPI

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to a quality assurance control method in a network.
[0002]
[Prior art]
In recent years, broadband access interfaces such as ADSL (Asymmetric Digital Subscriber Line) and FTTH (Fiber To The Home) have become widespread. Accordingly, via a network such as the Internet, a broadband information service represented by VOD (Video On Demand) or live broadcasting, or a two-way voice communication service represented by VoIP (Voice over Internet Protocol) is provided. The demand for is increasing. When a carrier, an ISP (Internet Services Provider), an IDC (Internet Data Center), or the like provides the above services, the following quality assurance is required.
(1) It is necessary to constantly allocate a band in End-to-End (end-to-end).
(2) The allocated bandwidth is kept used for a certain period of time, and must not be adversely affected by other traffic while using the allocated bandwidth.
(3) The assigned bandwidth differs depending on the service used by the network user. For this reason, it is necessary to flexibly assign the allocated bandwidth to the service content.
[0003]
In order to realize a broadband information service and a two-way voice communication service, it is necessary to perform the quality assurance shown in the above (1) to (3) in a network. The network may be closed by a carrier network. The prior art of the quality assurance method in the network is shown below.
[0004]
<DiffServ method>
The DiffServ method is an architecture for coordinating a plurality of network devices to maintain a high transfer quality of a flow. In the DiffServ method, flows are classified into several classes, there is a difference in transfer quality between them, and packet transfer is performed. This enables provision of relatively high-quality packet transfer. The DiffServ method has a mechanism in which a router at the entrance of a network writes a class identifier for classifying a packet into a packet, and a router inside the network performs packet transfer according to the prioritization based on the class identifier.
[0005]
<Flexible bandwidth control using ASIC technology>
Bandwidth control is a type of QoS (Quality of Service), and limits the available line capacity (communication speed) for each type of packet. It is a function of a network device (router) that constitutes a network, and is used to smoothly transmit and receive high-priority packets and to prevent loss of packets by adjusting the communication speed in advance.
[0006]
An ASIC (Application Specific Integrated Circuit) is an integrated circuit that is designed and manufactured for a specific application. Vendors that provide network devices (routers) use ASIC technology to realize a flexible bandwidth control function. Implements an integrated circuit using a network device.
[0007]
Flexible bandwidth control using ASIC technology is based on source or destination IP address, source or destination TCP / UDP port number and port identified by protocol type, packet priority, layer 4 session flow Realizes bandwidth control that can be provided at a reduced wire speed.
[0008]
<MPLS QoS path method>
MPLS (MultiProtocol Label Switching) is a packet transfer technique using a label switching method. When the MPLS function is implemented in a network device (router, called “Label Switching Router: LSR” in the case of a router or MPLS) constituting a network, in the MPLS network, the packet is forwarded to any router according to the destination of the packet. LSR holds such information, and each route is identified by a label. When a packet arrives at a network device (edge router) at the entrance of the network, a label is added to the route information in the packet, and the packet is transferred to the next router. The next network device (core router) determines which router to forward by looking at the label added to the packet, and sends the packet to an appropriate destination. The network device (edge router) at the exit to the external network removes the label from the arriving packet and transfers the packet to the external router. The MPLS QoS path method is a technique for securing a band in each network device (router) using a Resource reSerVation Protocol (RSVP) for a preset path (path) called an MPLS LSP (Label Switching Path). It is.
[0009]
When performing quality control in a network using the above-mentioned "DiffServ method", "flexible bandwidth control method using ASIC technology", or "MPLS QoS path method", a network device (router) configuring the network is required. On the other hand, it is necessary to make various settings in advance. Further, when quality assurance is performed by end-to-end, it is necessary to perform settings necessary for quality assurance for all network devices (routers) on a path (route). It is very difficult for a network operator to manually make settings for quality control for all network devices (routers). Therefore, there is a policy server method as a conventional technique for reducing the network operation cost.
[0010]
<Policy server method>
The policy server method is a method of collectively setting network-wide QoS settings for a plurality of network devices (routers) constituting a network, such as making an end-to-end quality assurance request (bandwidth reservation request). This is a method that uses a network operation management server that has functions such as performing. A policy is a setting guideline such as quality assurance (bandwidth reservation) and route selection for a network device (router). If a policy server is used, there is no need to perform quality assurance settings for each network device (router) that configures the network, and QoS such as end-to-end quality assurance (bandwidth reservation) and route selection is relatively easy. Guarantees can be made.
[0011]
<Quality assurance request reception method (admission control method)>
The quality assurance request reception method (admission control method) is used to provide a broadband information service represented by VOD or live broadcasting via a "policy server method" or a two-way voice communication service represented by VoIP or the like. In this method, the policy server receives an end-to-end quality assurance request from a network user, analyzes the quality assurance request from the network user, and performs quality assurance setting on a network device (router). .
[0012]
Further, there is a flow aggregation method as an extension of the quality assurance request reception method (admission control method). The flow aggregation method, such as “DiffServ method”, “flexible bandwidth control method using ASIC technology”, or “MPLS QoS path method”, guarantees the bandwidth for each end-to-end flow as quality control in the network. Instead of performing QoS guarantees such as, for example, a content provider or a network operator predicts network resources required for providing a service, and predicts the required network resources (for example, a bandwidth for a plurality of flows). ) Is reserved in advance.
[0013]
When the quality assurance request reception method (admission control method) is realized by using the flow aggregation method, provisioning of network resources predicted to be necessary is performed in advance using a policy server or the like. Thereafter, when a quality assurance request is received from a network user, it is determined whether there are any remaining network resources that can accept the request. If there are any remaining network resources, the service is provided to the network user. Is allowed.
[0014]
<Dynamic load balancing method using multipath>
As a non-guaranteed traffic transfer method, there are a shortest path-based hop-by-hop transfer method and a dynamic load distribution method using multipath. In the shortest path-based hop-by-hop transfer method, a shortest path is selected by a path search mechanism of an IGP (Interior Gateway Protocol) such as OSPF (Open Shortest Path First), and the selected shortest path is routed by a network device (router). Reflected in the table. The transferred packet is transferred to the next hop in each network device (router) while referring to the routing table in packet units. The dynamic load distribution method using a multipath is a method in which a plurality of paths (routes) are set by MPLS, and traffic is distributed and transferred to the set plurality of paths (routes).
[0015]
As prior art relating to a dynamic load distribution method using multipath, a path selection device and method for a communication network disclosed in Patent Document 1, a transmission path control device and a transmission path control method disclosed in Patent Document 2, and There is a recording medium on which a transmission path control program is recorded.
[0016]
As a prior art according to the present invention, there is a communication quality control system and a communication quality control method disclosed in Patent Document 3.
[0017]
[Patent Document 1]
JP 2001-144804 A
[Patent Document 2]
JP 2001-320420 A
[Patent Document 3]
JP 2002-374286 A
[0018]
[Problems to be solved by the invention]
When performing quality control in a network such as the “DiffServ method”, the “flexible bandwidth control method using ASIC technology”, and the “MPLS QoS path method”, the “policy server method” It is considered to realize quality assurance in a broadband information service represented by VOD, live broadcasting, and the like, and a two-way voice communication service represented by VoIP, etc. In this case, in the worst case, the content provider or the network operator can use the policy server to provide a network device (router) with individual policies (“DiffServ method”, “flexible using ASIC technology”) for each user. It is necessary to set quality control such as “a suitable bandwidth control method” and “MPLS QoS path method”. For this reason, a content provider or a network operator who operates the network on a daily basis needs skills for performing extremely complicated settings. Furthermore, even if the content provider or the network operator has high skills, the time required for setting cannot be ignored. Further, operation costs including labor costs and the like increase.
[0019]
When the quality assurance request reception method (admission control method) is used, quality control in a network such as “DiffServ method”, “flexible bandwidth control method using ASIC technology”, and “MPLS QoS path method” is performed by a policy server. Is performed for each End-to-End flow. For this reason, the control of the network device (router) increases, the processing performance of the policy server deteriorates, and the response to the quality assurance request to the network user deteriorates. Furthermore, since the network device (router) has an upper limit value for which quality control can be set for each end-to-end flow, the allowable number of services is limited.
[0020]
On the other hand, using a quality assurance request reception method (admission control method) using a flow aggregation method, cooperates with a broadband information service represented by VOD or live broadcasting, and a two-way voice communication service represented by VoIP, etc. Consider a case where quality assurance in an effective network is performed. In this case, the content provider or the network operator predicts the network resources required for providing the service, and secures the required network resources (for example, the bandwidth for a plurality of flows) in advance. It is necessary to keep. In addition, if a content provider or network operator accepts a request that exceeds the upper limit of network resources set and secured in advance, the service shall be permitted even if other paths (paths) are available. Can not do. For this reason, network resources cannot be used effectively.
[0021]
As described above, when trying to achieve quality assurance in a broadband information service typified by VOD or live broadcasting or a two-way voice communication service typified by VoIP, operation costs for quality assurance increase. I do. On the other hand, there is the following problem even if the quality assurance request of the network user is accepted and the operation cost is reduced.
(I) Without flow aggregation
-Since the setting for the network device (router) is performed for each quality assurance request, the response to the quality assurance request deteriorates.
-The allowable number of services depends on the upper limit that can be set for each flow of the network device (router).
(Ii) With flow aggregation
-It is necessary to predict the network resources required for providing the service and to secure the required network resources in advance.
-When a request that exceeds the upper limit of the network resources secured in advance is accepted, the service cannot be permitted even if other paths (paths) are available.
[0022]
Therefore, unless the problems described above are solved, quality assurance cannot be realized in a broadband information service represented by VOD or live broadcasting or a two-way voice communication service represented by VoIP. In addition, carriers, ISPs, IDCs, and the like cannot provide a service that can collect a service usage fee, which may be a major obstacle to business development.
[0023]
In addition, carriers, ISPs, and IDCs that provide broadband information services such as VOD and live broadcasting, and two-way voice communication services such as VoIP, not only consider traffic for quality assurance, Care must be taken to minimize non-guaranteed traffic as much as possible. In the “dynamic load distribution method using multipaths”, what ratio (distribution ratio) traffic flows to a plurality of paths (routes) is determined. For this reason, the traffic actually flowing in the network is measured, and the distribution ratio is determined so that the traffic flows efficiently by distributing the traffic on a plurality of paths (routes). However, if the “dynamic load distribution method using multipath” is applied together with the quality assurance request reception method with flow aggregation (admission control method), it is already secured (will be used in the future). 2.) The distribution ratio of traffic flowing through a plurality of paths (routes) needs to be determined in consideration of network resources.
[0024]
An object of the present invention is to provide a quality assurance control technique that can reduce the burden on a network service provider.
[0025]
Another object of the present invention is to provide a quality assurance control technique capable of improving the responsiveness to a use request.
[0026]
It is another object of the present invention to provide a load distribution technique capable of determining a distribution ratio for a path in consideration of quality assurance traffic.
[0027]
[Means for Solving the Problems]
The present invention employs the following configuration.
[0028]
That is, the first aspect of the present invention is provided in a network configured by network devices having a quality assurance function,
Use request frequency acquisition means for obtaining a use request frequency as a frequency of a use request for a network service including a quality assurance request from a user;
Control means for performing quality assurance network control based on the use request frequency,
And a quality assurance control device.
[0029]
According to the present invention, it is possible to perform network control such that provisioning for a use request is automatically performed according to the use request frequency.
[0030]
Preferably, the control means in the first aspect performs a quality assurance network control for securing a network resource for quality assurance capable of allocating another traffic flow together with a traffic flow for a certain use request, and The traffic flow corresponding to the use request is allocated to the network resource for quality assurance.
[0031]
Such quality assurance network control is configured to be performed, for example, when the frequency of use requests from users for network services is monitored, and it is determined that the use requests tend to occur continuously. be able to.
[0032]
Also, the frequency of usage requests from users for network services is monitored, and if it is determined that usage requests do not tend to occur continuously, quality assurance of only the traffic flows required by the services is guaranteed. It is also possible to configure so that the control means performs network control.
[0033]
Preferably, the control means in the first aspect performs quality assurance network control for reserving a network resource for quality assurance capable of accommodating a plurality of traffic flows, and reserves a traffic flow for a use request generated thereafter. The network resources for quality assurance.
[0034]
According to such a configuration, the frequency of use requests for network services from users is monitored, and a plurality of traffic flows are accommodated before receiving a use request for network services from users. It is possible to perform quality assurance network control that can be (aggregated).
[0035]
According to the above configuration, provisioning (for example, securing network resources) for the next use request is performed for a network resource for quality assurance (for example, quality assurance path (for example, quality assurance path) that can accommodate (aggregate) other traffic flows. This is automatically performed by securing a QoS path)) or by reserving network resources for quality assurance (prior reservation). Therefore, it is possible to improve the promptness of the response to the next use request. Further, the network service provider does not need to secure network resources for each use request. This reduces the burden on the provider.
[0036]
In the first aspect, it is preferable that the use request frequency is classified into a plurality of levels, and the control means performs the above-described quality assurance network control according to the current level. In this case, the number of levels prepared and the contents of the quality assurance network control performed according to the levels can be appropriately combined.
[0037]
Preferably, the control means according to the first aspect, when the traffic flow for the newly accepted use request cannot be accommodated in the already secured network resource for quality assurance, -Reserve new quality assurance network resources to which flows can be assigned.
[0038]
In this case, it may be configured to secure a network resource for quality assurance for allocating only a traffic flow for one use request. However, a configuration is adopted in which a new network resource for quality assurance that can allocate traffic flows for multiple usage requests is secured, and the traffic flow for the next usage request is allocated to the network resources for quality assurance. Is preferred.
[0039]
According to such a configuration, for example, when the allowable capacity of the service use request of the network resource secured by the quality assurance network control capable of accommodating (aggregating) a plurality of traffic flows is exceeded (for example, when the capacity is secured). (If the bandwidth exceeds the bandwidth of the network resource), it is searched for an alternative route for which quality assurance is possible, and if the alternative route is detected, the alternative route can be automatically provisioned. This allows effective use of network resources.
[0040]
Preferably, the control means in the first aspect maintains the network resource for quality assurance for a certain period of time even when there is no traffic flow accommodated in the quality assurance path, and newly uses the network resource during that time. When a request occurs, the traffic flow for the use request is allocated to the quality assurance network resource.
[0041]
According to such a configuration, for example, when the allowable capacity of the service use request capable of guaranteeing the quality of the network resource secured by the quality assurance network control capable of accommodating (aggregating) a plurality of traffic flows is exceeded (for example, , If the bandwidth exceeds the secured bandwidth), the newly secured network resource for quality assurance is marked as a frequently used network resource for quality assurance, and the marked network for quality assurance is marked. -Regarding resources, even if there are no users, they are treated as reserved network resources for quality assurance for a certain period of time. As a result, efficient network resource management can be performed.
[0042]
According to the above configuration, the traffic flow for the next use request can be allocated to the secured network resources for quality assurance. As a result, the promptness of the response to the use request can be improved. In addition, the service provider does not need to secure network resources for quality assurance for each use request.
[0043]
Preferably, the control means according to the first aspect monitors an available bandwidth of the network resource for quality assurance, and when the available bandwidth exceeds a predetermined threshold, the quality assurance newly secured. The setting value of the bandwidth of the network resource for communication.
[0044]
In such a configuration, the free bandwidth of the network resource for quality assurance is monitored, and if the network resource for quality assurance is not sufficiently used, the network resource for quality assurance set next is set. Resources are reserved with a smaller bandwidth than before. Thereby, effective use of the entire network resources can be achieved.
[0045]
A second aspect of the present invention is a load distribution device for dynamically distributing traffic flowing through a network to a plurality of routes,
Free bandwidth calculation means for calculating the free bandwidth of the link of each route;
Means for determining a distribution ratio of traffic to the plurality of routes based on an available bandwidth of a link of each route,
When a band for quality assurance is secured in the link, the free band calculation means subtracts the free band in the band for quality assurance and the actually used band of the link from the link band, and Find a free band.
[0046]
According to the second aspect, for example, when dynamic load balancing is applied to non-guaranteed traffic, traffic flowing through the network is used as information for determining a distribution ratio for distributing traffic flowing through the network to a plurality of routes (routes). In addition to the measured traffic, the remaining capacity of the service allowance that can be guaranteed and already secured (for example, the remaining bandwidth of the reserved bandwidth) is determined to be likely to be used immediately, and is added to the measured traffic. This can be used as the determination information of the dispersion ratio.
[0047]
According to the second aspect, in determining the traffic distribution ratio for a plurality of routes, the free bandwidth of the quality assurance bandwidth is taken into account in calculating the free bandwidth. As a result, for example, after the non-guaranteed traffic transferred by the best effort is set to a certain route, the quality assurance traffic is allocated to the free band of the band for the quality assurance, the bandwidth of the non-guaranteed traffic is narrowed, and the transmission speed is reduced. Can be suppressed.
[0048]
Preferably, the vacant band calculation means in the second aspect converts the vacant band in the band for quality assurance from the link band only when it is determined that there is a high possibility that the vacant band in the quality guaranteed path is used. Reduce.
[0049]
According to such a configuration, only when there is a high possibility that the adverse effect on the non-guaranteed traffic will occur, the calculation of the free bandwidth of the link in consideration of the free bandwidth in the bandwidth for quality assurance is performed.
[0050]
In this case, for example, when it is determined that there is a high possibility that the remaining capacity of the service allowable capacity that can guarantee the quality (for example, the remaining bandwidth of the secured bandwidth for quality assurance) is likely to be used immediately, Only the vacant bandwidth of the bandwidth for quality assurance, which is frequently used by the user for network services, is added to the actually measured traffic (actually used bandwidth) as predicted traffic, and can be used as the distribution ratio determination information. . This makes it possible to realize a dynamic load distribution method using a multipath that incorporates a more accurate predictive element.
[0051]
According to a third aspect of the present invention, a quality assurance control device provided in a network including network devices having a quality assurance function includes:
Obtain the usage request frequency as the frequency of usage requests for network services including quality assurance requests from users,
Perform quality assurance network control according to the usage request frequency
Quality assurance control method.
[0052]
According to a fourth aspect of the present invention, there is provided a load distribution device for dynamically distributing traffic flowing through a network to a plurality of routes,
When calculating the available bandwidth of the link of each route, if a bandwidth for quality assurance is secured for the target link, the available bandwidth in the bandwidth for quality assurance and the actually used bandwidth of the link are used. Subtract the link bandwidth to determine the free bandwidth of the link,
Determining a traffic distribution ratio for the plurality of routes based on an available bandwidth of a link of each route;
And a load distribution method.
[0053]
The present invention can also be specified as a program that causes a computer to function as the quality assurance control device or the load distribution device according to the present invention, or as a recording medium on which the program is recorded.
[0054]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, embodiments of the present invention will be described with reference to the drawings. The configuration in the embodiment is an exemplification, and the present invention is not limited to the configuration in the embodiment.
[0055]
〔Overview〕
First, an outline of an embodiment of the present invention will be described. FIG. 1 shows a schematic diagram of the prior art related to the present invention. FIG. 2 shows a schematic diagram of an embodiment of the present invention that solves the above-described problems.
[0056]
As shown in FIG. 1, the policy server is connected to a network including network devices (routers) having a quality control function. Each network device has a quality assurance function to which a "Diffserv method", a "flexible bandwidth control method using ASIC technology", and an "MPLS QoS path method" are applied.
[0057]
The policy server is configured to implement the policy server method. The policy creation unit creates a policy for a network device, the policy creation GUI unit provides a GUI (Graphical User Interface) for creating a policy, and the policy creation unit creates a policy. It comprises a policy management control unit for managing and controlling, a policy application unit for applying the policy to each network device, and a policy management DB for storing information on the policy. In addition, the policy server is configured to implement a quality request reception method, and a quality assurance request receiving means for receiving a quality assurance request from a network user, and an admission for securing network resources according to the quality assurance request. It has a control means and a network resource management DB for storing information such as the usage status of network resources.
[0058]
Also, as shown in FIG. 1, a server (load distribution server) that implements a dynamic load distribution method using multipath is connected to a network. The load distribution server includes a network traffic measurement unit that measures a traffic amount of each path (path) in the network, a load distribution calculation unit that calculates a traffic distribution ratio for each path based on the traffic amount of each path, Load balancing control means for performing a traffic distribution process so that traffic for a route falls within a range of an appropriate distribution ratio; and network statistical information for storing network-related statistical information such as a traffic amount of each route measured by the measuring means. DB. The “dynamic load distribution method using multipath” can be implemented by a server alone, a network device alone, or a server and a network device together.
[0059]
In FIG. 2, a network 100 includes a plurality of network devices (routers) 101 to 104. However, the number of network devices (routers) depends on the scale of the network. The network 100 has a network configuration in which a plurality of paths (routes: also referred to as “routes”) exist. For example, the route of the network 100 between the client 105 and the server 106 includes a route of the router 102-the router 101-the router 104 and a route of the router 102-the router 103-the router 104. Each route is composed of nodes (network devices) and links (physical lines) connecting the nodes. The network 100 has a configuration in which a QoS path (LSP) can be set on an end-to-end route in the network 100.
[0060]
The network devices (routers) 101 to 104 have a quality assurance function for realizing a quality assurance method such as a "DiffServ method", a "flexible bandwidth guarantee method using ASIC technology", and an "MPLS QoS path method".
[0061]
The client 105 is a machine (terminal device) used by a network user, and is connected to the network 100.
[0062]
The server 106 is a device that provides a broadband information service typified by VOD, live broadcasting, and the like, and a two-way voice communication service typified by VoIP and the like, and is connected to the network 100.
[0063]
The quality assurance server 107 is a control device (quality assurance control device) having the quality assurance control method according to the present invention, and is connected to the network 100. The quality assurance server 107 includes the following means and a DB (Database). The quality assurance server 107 can be configured using a computer (information processing device) having a communication function with a network device such as a dedicated server machine, a personal computer, or a workstation. The computer constituting the quality assurance server 107 includes a processor such as a CPU and a recording medium including a main storage and an auxiliary storage (recording programs and data). The processor stores various programs stored in the auxiliary storage. By loading the program into the main memory and executing it, it functions as the quality assurance server 107 having each unit (each unit) and a database (DB) as shown in FIG. In the following description of each means and DB (component), “(remodeling)” indicates a component that can be realized by remodeling an existing component, and “(new)” indicates This indicates a new component for realizing the invention, and “(existing)” indicates an existing component.
[0064]
-Quality assurance request receiving means 200 (modification)
The quality assurance request receiving unit 200 is a communication unit that receives a quality assurance request from the client 105.
[0065]
-Quality assurance request analysis means 201 (new)
When the QoS path request type of the quality assurance parameter is “set”, the quality assurance request analysis unit 201 performs a time stamp process at the time of the request.
[0066]
-Usage status storage means 202 (new)
When the quality assurance request is received from the network user, the usage status storage unit 202 registers the service usage frequency information in the service usage status management DB 301 and updates the service usage frequency information. The service use frequency can be managed for each content, server, and network (segment) that the network user is trying to use. In addition, the service use status management DB 301 is updated.
[0067]
・ Usage status judgment means 203 (new)
The usage status determination unit 203 calculates a service usage frequency per time.
[0068]
-QoS path setting means 204 (new)
First, the QoS path setting unit 204 checks with the reception control unit 205 whether or not there is already a path (route) with “flow aggregation”. If there is already a path (route) with “flow aggregation”, the service permission is returned to the network user.
[0069]
On the other hand, if there is no already secured path (route) with “flow aggregation”, the QoS path setting unit 204 sends a request to the route search unit 206 to search for another path (route). Search for an alternative route. If no alternative route is found, the QoS path setting means 204 returns service rejection to the network user.
[0070]
On the other hand, if an alternative route is found, the QoS path setting unit 204 requests the QoS path management unit 207 to control the network device (router) and stores the QoS path setting information in the network resource management DB 302. Set. Then, when the control of the network device is completed, the QoS path control means 204 returns a service permission to the service user.
[0071]
・ Reception control means 205 (new)
The reception control unit 205 checks whether the path (path) specified by the reception control parameter is already registered in the network / resource management DB 302 as a path (path) with “flow aggregation”.
[0072]
-Route search means 206 (modification)
The route search unit 206 searches for a specified end-to-end path (route).
[0073]
-QoS path management means 207 (new)
The QoS path management unit 207 requests the network control unit 208 to control the network device (router) and registers the newly secured path (route) in the network resource management DB 302.
[0074]
・ Network control means 208 (modification)
The network control unit 208 analyzes the band setting parameters and controls network devices (routers).
[0075]
-Flow aggregation marking means 209 (new)
The flow aggregation marking unit 209 registers a flag indicating that the reserved band has been exceeded in the network / resource management DB 301 in the processing by the reception control unit 205.
[0076]
-Dijkstra calculation means 210 (remodeling)
The Dijkstra calculation means 210 performs Dijkstra calculation (search for the shortest path by the Dijkstra method).
[0077]
-QoS path cycle management means 211 (new)
The QoS path cycle management unit 211 starts periodically and manages a reservation type QoS path.
[0078]
-QoS path cycle release means 212 (new)
The QoS path period releasing unit periodically starts and releases the QoS path if there is QoS path setting information stored in the network resource management DB 301 in which a flag indicating that the reserved band has been exceeded is set. .
[0079]
-QoS path aggregation flow management means 213 (new)
The QoS path aggregation flow management unit 213 periodically checks whether or not the QoS path aggregation flow has secured an appropriate reserved bandwidth.
[0080]
・ Load distribution calculation means 214 (modification)
The load distribution calculation unit 214 is started periodically, checks whether or not to perform load distribution processing, and performs load distribution control processing when it is necessary to perform load distribution.
[0081]
・ Network traffic measurement means 215 (existing)
The network traffic measuring unit 215 measures the traffic flowing to the network device (router) via SNMP (Simple Network Management Protocol) and registers the traffic in the network statistical information DB 303.
[0082]
-Load distribution control means 216 (existing)
The load distribution control unit 216 performs traffic load distribution control on network devices (routers).
[0083]
-Service usage status management DB 301
The service usage status management DB 301 is a database that stores and manages information related to service usage frequency in units of content, server, and network (segment) that a network user intends to use.
[0084]
・ Network resource management DB 302
The network resource management DB 302 is a database that stores and manages information related to a network path (route).
[0085]
・ Network statistics information DB303
The network statistical information DB 303 is a database that stores and manages network statistical information (such as the traffic volume of each path).
[0086]
FIG. 3 is an explanatory diagram of a provisioning method according to a service use status, and FIG. 4 is an explanatory diagram of a flow reuse type provisioning method according to a service use status. In the present invention, there is provided a usage status management (corresponding to the usage status storage unit 202 and the usage status determination unit 203) that manages content units, server units, and network units (segment units) that the network user intends to use. . This makes it possible to perform provisioning according to the service usage frequency per time (corresponding to the QoS path management means 207 and the QoS path cycle management means 211).
[0087]
More specifically, a plurality of levels, for example, levels 1 to 4 as shown in FIG. 3 can be prepared according to the usage status (use request frequency) per hour for the service. In this example, the higher the level value, the higher the usage request frequency. Here, level 1 is “without on-demand type flow aggregation”, level 2 is “with on-demand type flow aggregation”, level 3 is “without reservation type flow aggregation”, and level 4 is “reservation type flow aggregation”. There is flow aggregation ". Then, processing for securing a traffic flow (hereinafter, referred to as “flow”) according to the current level value is performed.
[0088]
The usage status management detects the current level value by inquiring of the usage status. At this time, if the level value is level 1, the QoS path (quality assurance path: network for quality assurance) in which only the bandwidth of the flow (on-demand type flow) for the user's current use request is secured. Resource) is set in the QoS path setting. On the other hand, if the current level value is level 2, the usage status management sets the QoS path having the bandwidth that can aggregate other flows together with the flow for the usage request currently requested by the user to the QoS level. Set to path setting. If the level value is level 3, the usage status management sets the QoS path of the reserved flow, in which only the bandwidth of the flow for which the usage request from the user is predicted is secured, to the QoS path setting (period). . On the other hand, if the level value is level 4, the usage status management causes the QoS path of the reserved flow in which the bandwidth that can aggregate a plurality of flows is secured to be set to the QoS path setting (period).
[0089]
As described above, according to the embodiment of the present invention, when the usage frequency tends to increase, a QoS path that can aggregate a plurality of flows or a reservation-type QoS path is set, and the flow according to the usage request is established. Can be set in the QoS path. As a result, a quick response (improvement in promptness) to the use request is possible.
[0090]
In addition, it is considered that an on-demand type QoS path secured by “with flow aggregation” is frequently used (frequently used). For this reason, as shown in FIG. 4, when a use request exceeding the flow aggregation is received at least once at level 2, even if there is no user using the QoS path of the flow aggregation, the QoS path is maintained for a certain period of time. Can be configured so as to retain the network resources for which the QoS path is set without releasing the network resources.
[0091]
As a result, when a use request is subsequently received, if there is an empty QoS path, it is possible to respond to the use request using this QoS path. Therefore, the responsiveness to the use request is improved.
[0092]
FIG. 5 is a diagram showing a relationship between a QoS path (GS (Guaranteed Services) flow) and non-guaranteed traffic (BES (Best Effort Services) flow). FIG. 6 is a diagram showing a level transition of the QoS path. FIG. 7 is a diagram illustrating a link having a plurality of aggregation paths.
[0093]
The GS flow is a flow for Guaranteed Services in which a QoS path for performing quality assurance described in the present invention is secured and a network is used. The BES flow is a flow for Best Effort Services that uses a network without guarantee.
[0094]
In the dynamic load balancing method using the multipath according to the conventional technology, when the GS flow and the BES flow are mixed, as shown in FIG. 5, the actually used bandwidth where the GS flow and the BES flow are combined ((2) in FIG. 5). ▼) is subtracted from the link bandwidth to make it an available bandwidth, and load balancing is performed so that the available bandwidth of the multipath is averaged. In the definition of the free band shown in FIG. 5, the band (1) in FIG. 5 that is not reserved for GS is also regarded as the used band. That is, even if the GS flow (GS traffic) is not actually flowing, it is counted as the used band.
[0095]
On the other hand, by performing load distribution in consideration of the band that will be used in the future and the unreserved band in the reserved band of GS, effective load distribution can be achieved when GS flows and BES flows are mixed. And Further, among the GS flows, there is a GS flow that is not used much (actually, the BES flow allows more effective utilization of network resources because the GS flow has a vacant time before it is used). For this reason, it is not appropriate to consider all GS flows when performing dynamic load balancing using multipath.
[0096]
Therefore, in the present invention, in order to perform more appropriate prediction, a free band calculation method according to the level of the QoS path is performed. The level transition of the QoS path can be considered in cases A, B, and C as shown in FIG.
[0097]
Case A is a case in which a QoS path (level 1 path) is set for each GS flow in level 1 (initial state) (A in FIG. 6). Case B uses a QoS path in which a plurality of GS flows are aggregated (level 2 aggregated path) in level 2 (a state of high use frequency), and the traffic of the GS flow is set. This is the case that is flowing (B in FIG. 6). Case C is a case in which at level 1 (infrequently used), a QoS path (aggregation path of level 2) for aggregating flows has free space, but the number of GS flows accommodated is reduced (FIG. 6C).
[0098]
In case B, the frequency of use is high, and there is a possibility that the unreserved band (1 in FIG. 6: free band in the band for quality assurance) in the level 2 aggregated path shown in FIG. 6 is used immediately. high. For this reason, it is considered to be predicted traffic to be considered when calculating the available bandwidth in the dynamic load distribution method using multipath. On the other hand, in case A, since there is no reserved bandwidth due to the aggregated flow (reserved QoS path), it is not necessary to consider the free bandwidth calculation in the dynamic load distribution method using multipath. In case C, the usage frequency is low, and it is considered that the GS flow in the aggregated path is released with a lapse of time and after a certain period of time. Therefore, in the calculation of the available bandwidth in the dynamic load distribution method using the multipath, the predicted traffic that is highly likely to use the unsecured bandwidth ((2) in FIG. 6) in the aggregated path of level 2 immediately is used. Thinking is not considered appropriate. Therefore, in the present invention, only the aggregated flow of case B is applied as the predicted traffic to be considered when calculating the available bandwidth in the dynamic load distribution method using multipath.
[0099]
In addition, as shown in FIG. 7, even when a plurality of aggregated flows exist in one link, only the aggregated flow of case B is considered. The present invention includes a mechanism for periodically checking whether or not the aggregated flow of the QoS path is appropriately assigned.
[0100]
In the present invention, as means for considering only the aggregated flow of case B, the network traffic measuring unit 209, the load distribution calculating unit 214, the load distribution control unit 216, and the QoS path aggregated flow monitoring unit 213 realize the operation.
[0101]
〔Example〕
Examples of the present invention will be described below. FIG. 8 is a system configuration diagram for explaining an embodiment of the present invention. In the embodiment, a description will be given using a VOD service which is a broadband information service as an example.
[0102]
In FIG. 8, a network 100 in which a carrier, an ISP, an IDC, and the like provide a VOD service includes a network device A having “172.27.1.1” as a representative address (an IP address for specifying the network device), and a representative address. Network device B having “172.27.2.1” as the representative address, network device C having “172.27.3.1” as the representative address, and “172.27.4.1” as the representative address. It is configured with a network device D.
[0103]
The network device A is connected to the network device B (connected by the interface of “172.27.10.1” of the network device A) and the network device D (“172.27.40.2” of the network device A). Connection with the interface). The network device B is connected to the network device A (connected by the interface 172.27.10.2 of the network device B) and the network device C (connected by the interface 172.27.20.2 of the network device B). Connection) with a link. The network device C is connected to a network device B (connected by an interface of “172.220.2” of the network device C) and a network device D (interface of “172.27.30.1” of the network device C). With a link). The network device D is connected to the network device A (connected by the interface “172.27.40.1” of the network device D) and the network device C (interface “172.27.30.2” of the network device D). Connection) and have a link.
[0104]
Also, a quality assurance server 107 (IP address: 172.27.5.100) provided with various units for realizing various means (FIG. 2) according to the present invention, and a service reception for receiving a user request for the VOD service The server 108 (IP address: 172.27.5.200) is connected to the network device A, and as servers for providing a VOD service, a VOD server 106A (IP address: 172.27.7.100) and a VOD server 106B (IP address: 172.27.6.100) is connected to the network.
[0105]
The VOD server 106A is connected to the network device C, and the VOD server 106B is connected to the network device D. Further, the user who uses the VOD service performs the operation via the client 105 (IP address: 172.27.8.100) connected to the network device B.
[0106]
A user who intends to use the VOD service operates the client 105 to access the service reception server 108. Then, on the display of the client 108 used by the user, a login screen 14 (see the display example in FIG. 9) for performing login is displayed.
[0107]
The user inputs a user ID and a password on the login screen 14 and clicks a login button. Then, an authentication process is performed on the service reception server 108, and when login is permitted, a content list screen 15 (FIG. 10) for selecting a content desired by the user and a quality assurance level is displayed on the display of the client 105. Display example) is displayed.
[0108]
The user selects a desired content and a quality assurance level using the content list screen 15, and clicks a request button. Then, information indicating the desired content and the quality assurance level is transmitted to the service reception server 108.
[0109]
The service reception server 108 checks whether or not quality assurance is necessary when referring to the content from the user (when using the VOD service). Send it to server 107.
[0110]
The quality assurance server 107 performs a quality assurance control process desired by the user, and returns the result to the service reception server 108. When the quality assurance control process is completed normally, the content is transferred from the VOD server (for example, the VOD server 106B) to the client 105 at the quality desired by the user (transfer rate: transmission band) and displayed on the display of the client 105. The selected content is reproduced on the displayed content reproduction screen 16 (see the display example in FIG. 11).
[0111]
FIG. 12 is a flowchart of the quality assurance control process from the login of the user to the reproduction of the content. FIG. 13 is a diagram illustrating the contents of a login message transmitted to the service reception server 108 when a user logs in. FIG. 14 is a diagram showing the contents of user information, which is data used when the service reception server 108 performs authentication. FIG. 15 is a diagram illustrating the content of the content information referred to when the service reception server 108 displays the content list. FIG. 16 is a diagram showing the content of the content selection message transmitted to the service reception server 108 when the user selects the content or the quality assurance level. FIG. 17 is a diagram showing the contents of a quality assurance message sent to the quality assurance server 107 when the service reception server 108 determines that quality assurance is required. FIG. 18 is a diagram showing the contents of a guarantee response message as a result of the quality assurance control process. FIG. 19 is a diagram illustrating the content of a content selection response message returned from the service reception server 108 to the client 105 after receiving the result of the quality assurance control process.
[0112]
In FIG. 12, when the user (user) inputs a user ID and a password via the login screen 14 and clicks the login button, the user ID, the password, and the user IP address (the IP address of the user terminal (client)) are included. The login message 31 (FIG. 13) is transmitted from the client 105 to the service reception server 108 via the network 100 (S1).
[0113]
Upon receiving the login message 31, the service reception server 108 acquires the user information 32 (user ID and password: FIG. 14) for performing the authentication process, and obtains the user information 32 and the user included in the login message 31. The ID and the password are collated to determine whether the user name and the password are correct (S2). At this time, if the user name and the password are incorrect (S2; NO), the service receiving server 108 transmits an NG message to the client 105.
[0114]
On the other hand, if the user name and the password are correct, the service reception server 12 collects and refers to the content information 33 (FIG. 15) (S3), transmits the content information 33 to the client 105, and displays the content selection screen ( For example, the client 105 displays the content list screen 15 (FIG. 10) (S4).
[0115]
When the user inputs necessary items on the content selection screen, a content selection message 34 (FIG. 16) is transmitted from the client 105 to the service reception server 108 via the network 100 (S5). As shown in FIG. 16, the content selection message 34 can include a content ID, a QoS path request type, a transfer rate, a user IP address, and a guarantee flag.
[0116]
The service reception server 108 determines whether the guarantee flag included in the content selection message 34 from the client 105 has a guarantee (on) or no guarantee (off) (S6). At this time, if the guarantee flag is on (S6; YES), the quality assurance request message 35 (FIG. 17) is transmitted from the service reception server 108 to the quality assurance server 107 via the network 100. As shown in FIG. 17, the quality assurance request message 35 includes a content ID, a QoS path request type, a transfer rate, and a user IP address included in the content selection message.
[0117]
Upon receiving the quality assurance request message, the quality assurance request receiving means (quality assurance request receiving unit) 200 of the quality assurance server 107 performs quality assurance control processing (S8; described later), and the quality assurance response message 36 according to the result. (FIG. 18) is transmitted to the service reception server 108 via the network 100 (S9). As shown in FIG. 18, the quality assurance response message 36 includes the response result (OK / NG) of the content selection and the IP address of the selected server.
[0118]
Upon receiving the quality assurance response message 36, the service reception server 108 transmits a content selection response message 37 (FIG. 19) including the content (response result, service IP address) included in the message to the client 105 via the network 100. (S10).
[0119]
Then, an application for providing the content to the user is activated in the client 105, and the content is provided to the user (client 105) from the VOD server 106A and / or 106B via the network 100 in a state where the quality is guaranteed ( S11). For example, the video content selected by the user is downloaded and reproduced in the band requested by the user.
[0120]
FIG. 20 is a flowchart showing the processing of the quality assurance request receiving means 200 of the quality assurance server 107. 20, the quality assurance request receiving means 200 waits for reception of a quality assurance request message 35 from the service reception server 12 (S20001).
[0121]
When the quality assurance request message 35 is received (S20001: YES), the quality assurance reception unit 200 activates the quality assurance request analysis unit (quality assurance request analysis unit) 201 and passes the processing (S20002, S20003). At this time, the quality assurance request message 35 is passed to the quality assurance request analyzer 201 as a quality assurance parameter 38 (FIG. 21).
[0122]
As shown in FIG. 21, the quality assurance parameters 38 include a content ID, a QoS path request type, a transfer rate (bandwidth), and a user IP address. The content ID is an identifier for identifying the content selected by the user. The QoS path request identifier is an identifier indicating whether to request (set) or cancel the quality assurance. The transfer rate is data obtained by converting a quality level (a service level represented by “Gold” or “Sliver”) selected by a user into a QoS parameter such as a band. The user IP address is an IP address of a client necessary for performing End-To-End quality control. In this embodiment, the IP address “172.27.8.100” of the client 13 is set.
[0123]
After that, the quality assurance request receiving means 200 receives the processing result from the quality assurance request analyzing means 201, returns it to the service receiving server 108 (S20004), and ends the processing.
[0124]
FIG. 22 is a flowchart showing the processing of the quality assurance request analysis means 201. In FIG. 22, the quality assurance request analysis unit 201 performs a determination process of the QoS path request type included in the quality assurance parameter 38 received from the quality assurance request reception unit 200 (S20101).
[0125]
At this time, if the QoS path request type of the quality assurance parameter 38 is “setting” (S20101; YES), the quality assurance request analysis unit 201 sets the time when the quality assurance request reception unit 201 starts up by using the system function. (S20102). Then, the quality assurance request analysis unit 201 passes the usage status storage parameter 39 (FIG. 23) including the extracted time data as the request time to the usage status storage unit (use status storage unit) 202. As a result, the process is passed to the use status storage unit 202 (S20103).
[0126]
As shown in FIG. 23, the usage status accumulation parameter 39 includes a content ID, a QoS path request type (setting), a transfer rate, a user IP address, and a request time. After that, the quality assurance request analysis unit 201 receives the processing result from the use status storage unit 202, returns it to the quality assurance request reception unit 200 (S20104), and ends the process.
[0127]
If the QoS path request type of the quality assurance parameter is “released” in S20101 (S20101; NO), the quality assurance request analysis unit 201 transfers the processing to the QoS path setting unit (QoS path setting unit) 204 (S20105). . At this time, the quality assurance request analysis means 201 gives the QoS path setting release parameter 40 (FIG. 24) to the QoS path setting means 204 as an input parameter.
[0128]
As shown in FIG. 24, the QoS path setting release parameter 40 can include a content ID, a QoS path request type (setting / release), a transfer rate, a user IP address, and a use frequency level value. However, the use frequency level value is included in the QoS path setting release parameter only when the QoS path request type is “set”, and is not included when the QoS path request type is “release”.
[0129]
After that, the quality assurance request analysis unit 201 receives the processing result from the QoS path setting unit 204, returns it to the quality assurance request reception unit 200 (S20106), and ends the processing.
[0130]
FIG. 25 is a flowchart showing the processing of the usage status accumulation means 202. In FIG. 25, the usage status storage unit 202 is activated when the QoS path request type of the quality assurance parameter is “setting”. That is, the usage status storage unit 202 is activated by receiving the usage status storage parameter 39 (FIG. 23) including the QoS path request type of “setting” from the quality assurance request analysis unit 201.
[0131]
The usage status storage unit 202 stores the time at which the quality control request was received (request time in the input parameters) in the service usage status management DB 301 (FIG. 26) for each content ID (S20201). Then, the usage status storage unit 202 passes the process to the usage status determination unit (use status determination unit) 203 (S20202). At this time, the usage status storage unit 202 passes the usage status determination request parameter 41 (FIG. 27) to the usage status determination unit 203 as an input parameter.
[0132]
As shown in FIG. 26, the service use status management DB 301 has a request time queue capable of storing a plurality of (at least the number of status determination queues) request times of a quality assurance request for each content ID (content). I have. The request time included in the usage status accumulation parameter is stored in a predetermined position (latest position) of the request time queue.
[0133]
In the service usage management DB 301, usage frequency levels, usage frequency thresholds that are boundaries between the levels, and the number of status determination queues are registered. The number of status determination queues is a number for extracting the latest data from the request time queue when calculating the use frequency. The number of levels, the threshold of the use frequency, and the number of status determination queues can be appropriately set by the service provider or the like (carrier, ISP, IDC, etc.).
[0134]
Also, as shown in FIG. 27, the usage status determination request parameter 41 includes a content ID, a QoS path request type (setting), a transfer rate, and a user IP address.
[0135]
FIG. 28 is a flowchart showing the processing of the usage status determining means 203. The use status determination unit 203 analyzes the use frequency of each content. First, the usage status determination unit 203 extracts the number of status determination queues from the service usage status management DB 301 (S20301).
[0136]
Next, the usage status determination unit 203 searches the service usage status management DB 301 with the content ID included in the usage status determination request parameter 41 from the usage status storage unit 202, and stores the request time queue of the corresponding content ID in the status determination. Extraction is performed in ascending order of the number of queues (S20302).
[0137]
Next, the usage status determination unit 203 calculates a usage frequency value from the head time and the tail time of the extracted data (S20303). The use frequency value is obtained by the following formula.
[0138]
(Number of status determination queues) / (extracted latest request time)-(oldest request time in extracted queues)
For example, when the extracted latest request time is 10:10, the oldest request time in the extracted queue is 10:05, and the number of extracted queues is 10, the usage frequency is 2 [request number. / Min]. In this way, the use status determination unit 203 can calculate the number of use requests within a certain time as a use frequency value.
[0139]
Next, the use status determining unit 203 obtains a level (a level indicating the use frequency) corresponding to the use frequency value. That is, the usage status determination unit 203 refers to the service usage status management DB 301 and converts the usage frequency value to a usage frequency level value (S20304). Specifically, the usage status determining unit 203 compares the threshold value of the usage frequency registered in the service usage status management DB 301 with the calculated usage frequency value, and extracts a corresponding level value.
[0140]
For example, if the calculated usage frequency value is less than the level 1 threshold, level 1 is calculated as the level value, and if it is greater than or equal to the level 1 threshold and less than the level 2 threshold, level 2 is calculated as the level value. When level 3 or higher is set, level 3 and subsequent levels are determined in the same manner as the level 2 determination method.
[0141]
Next, the use state determination unit 203 passes the process to the QoS path setting unit (QoS path setting unit) 204 using the QoS path setting release parameter 40 (FIG. 24) including the use frequency level value as an input parameter (S20305). . As described above, the QoS path setting release parameter 40 includes the use frequency level value only at the time of “setting”.
[0142]
FIG. 29 is a flowchart showing the processing of the QoS path setting means 204 when the QoS path request type of the QoS path setting release parameter 40 is “setting”. In FIG. 29, the QoS path setting unit 204 first checks whether a QoS path of “with flow aggregation” that can be used in response to a use request from a user is already registered or not. The reception control parameter 42 (FIG. 30) is created from the setting release parameter 40, the reception control parameter 42 is set as an input parameter, and the process is passed to the reception control unit 205 (S20401).
[0143]
As shown in FIG. 30, the reception control parameter 42 includes a user IP address, a reserved band (a bandwidth corresponding to a transfer rate desired by the user), and a content ID.
[0144]
If the response result in the processing result (the reception control response parameter 43 in FIG. 31) returned from the reception control unit (reception control unit) 205 is “OK” (S20402; YES), the QoS path setting unit 204 executes the service. Returns a value indicating permission. Here, the reception control response parameter 43 whose response result is “OK” indicates that a QoS path capable of accommodating the flow according to the use request has already been set, and that the use request can be used to respond to the use request. .
[0145]
As shown in FIG. 31, the admission control response parameter 43 includes a response result (OK / NG) of the admission request, a user edge IP address, a server edge IP address, a path edge IP address list, and a server IP address. including. The user edge IP address is an IP address of an edge node (network device) accommodating a user terminal (client). In this example, the IP address of the network device B is set. The server edge IP address is an IP address of an edge node (network device) that accommodates a server that provides content. In this example, the IP address of the network device C or D that accommodates the VOD server 106A or 106B is set. Is done. The route edge IP address list is a list of edge IP addresses on the route selected as the optimal route, and the server IP address is the IP address of the server (VOD server 106A or 106B) that provides the content.
[0146]
If the response result in the reception control response parameter 43 is not “OK” (S20402; NO), the QoS path setting means 204 executes the cache server information 44 (FIG. 33) stored in the network / resource management DB 302 (FIG. 32). ) And the edge information 45 (FIG. 34) (S20403), create a route search parameter 46 as shown in FIG. 35, set the route search parameter 46 as an input parameter, and use a route search means (route selection). Unit) 206 to pass the process (S20404).
[0147]
This is because there is no QoS path capable of accommodating the flow corresponding to the use request, or the existing QoS path has no capability of accommodating the flow (exceeding the permissible capability). This is for setting an alternative route (new QoS path) on the network 100.
[0148]
Here, as shown in FIG. 32, the network resource management DB 302 includes cache server information 44 as information relating to a cache server that provides a VOD service (in this example, VOD servers 106A and 106B), and an edge node in the network. , And QoS path setting information 50 as information relating to QoS path setting, and QoS path aggregation method information 51 as setting method information relating to QoS path setting.
[0149]
As shown in FIG. 33, the cache server information 44 of the network resource management DB 302 includes, as shown in FIG. 33, a content ID which is an identification ID unique to the content, a server IP address which is an IP address of the cache server, and an edge node in which the cache server is accommodated. And a server edge IP address.
[0150]
As shown in FIG. 34, the edge information 45 of the network resource management DB 302 includes an edge IP address indicating an IP address of an edge node (label switching router (LSR)) and a segment (network) accommodated in the LSR. It includes a network address indicating a network address and a subnet mask indicating a subnet mask of the network address.
[0151]
Further, as shown in FIG. 35, the route selection (route search) parameter 46 includes a content ID, a cache server IP address list, a user edge IP address, a reserved bandwidth, and whether or not a constraint condition regarding the cache server can be applied. Number of routes to be calculated, applicability of maximum allowable hop number limit, maximum allowable hop number, applicability of non-passing node constraint, applicability of non-passing link constraint, non-passing node information, non-passing link information And whether or not a passing node constraint can be applied, and a passing node list.
[0152]
If the response result in the processing result (FIG. 36: route search response parameter 47) returned from the route searching unit 206 is not “OK” (NG) (S20405; NO), the QoS path setting unit 204 guarantees the quality. Since there is no QoS path that can perform the service, a value indicating service non-permission is returned to the use status determination unit 203 (S20405).
[0153]
On the other hand, if the processing result in the route search response parameter 47 is “OK” (S20505; YES), the QoS path setting unit 204 creates the QoS path management parameter 48 (FIG. 37) from the route search response parameter 47. Then, the QoS path management parameter 48 is set as an input parameter, and the process is passed to the QoS path management means 207 (S20406).
[0154]
As shown in FIG. 36, the route search response parameter 47 includes a response result (OK / NG) of the route selection request, a user edge IP address, a server edge IP address, a route edge IP address list, and a server IP address. , Reserved bandwidth.
[0155]
The QoS path management parameters 48 include, as shown in FIG. 37, the hop count, the user edge IP address, the server edge IP address, the route edge IP address, the server IP address, the reserved bandwidth, the content ID, and the like. , A pass trigger, a use frequency level, and a sequence number.
[0156]
The QoS path management processing unit 207 performs End-To-End QoS path setting for a network including the actual network devices A, B, C, and D, and sets the usage frequency level. Different QoS paths are set in accordance with.
[0157]
If the processing result (FIG. 38: QoS path management response parameter 49) returned from the QoS path management unit 207 is not “OK”, the QoS path setting unit 204 returns a value (“NG”) indicating that the service is not permitted. (S20407). On the other hand, if the processing result from the QoS path management unit 207 is “OK”, a value indicating service permission is returned (S20408).
[0158]
As shown in FIG. 38, the QoS path management response parameter includes a response result (OK / NG) of the route registration request.
[0159]
FIG. 39 is a flowchart showing the processing of the reception control means (reception control unit) 205. In FIG. 39, the reception control unit 205 refers to the used bandwidth of the QoS path setting information 50 (FIG. 40) stored in the network resource management DB 302 (S20501), and determines whether there is a free bandwidth in the request. (S20502; YES), and if there is a vacancy (S20501; YES), the used bandwidth of the QoS path setting information 50 stored in the network / resource management DB 302 is added to the requested bandwidth, and the determination result is returned (S20503). ).
[0160]
The QoS path setting information 50 includes, as shown in FIG. 40, a hop count, a user edge IP address, a server edge IP address, a route edge IP address list, a server IP address, a reserved band, a used band, It includes a content ID, a pass trigger, a sequence number, a time stamp, a reserved band over flag, a use frequency level, and in-use user information. The in-use user information includes, for each user, a user IP address and , Used band. Note that the QoS path setting information 50 is prepared and managed for each content and each QoS path.
[0161]
FIG. 41 is a flowchart showing the processing of the route search means (route search unit) 206. The route search unit 206 extracts the edge information 45 (FIG. 34) by referring to the cache server information 44 (FIG. 33) stored in the network / resource management DB 302 (S20601). Next, the route search unit 206 determines whether the extracted edge node is specified as a passing node (S20602). If it is specified, the process is passed to Dijkstra calculation means (Dijkstra calculation unit) 210 that performs the shortest route search (S20603).
[0162]
Subsequently, the route search unit 206 determines the return value (determination result) of the Dijkstra calculation unit (S20604). At this time, if the determination result is NG (S20604; YES), NG is returned to the route searching means 206, and the process ends. On the other hand, if the determination result is OK (S20604; NO), the route information between the user edge and the passing edge node is stored (S20605). Then, in order to perform the Dijkstra calculation again, the process is passed to the Dijkstra calculation means 210 (S20606).
[0163]
Subsequently, the route search unit 206 determines the return value of the Dijkstra calculation unit (S20607). If the determination result is NG (S20607; NO), the route search means 206 skips the process, and if OK (S20607; YES), obtains the cost of the cache server (S20608).
[0164]
Next, the route search unit 206 determines whether the total value of the link cost and the cache server cost is smaller than the previously calculated cost value (S20609). At this time, if the total value is larger than the cost value (S20609; NO), the route search means 206 skips the process, and if smaller (S20609; YES), replaces the route information with the previously obtained route information (S20610).
[0165]
Next, the route search unit 206 determines whether or not the processing by the Dijkstra calculation unit 210 has been accommodated for all the cache servers (S20611). At this time, if there are remaining cache servers, the processing is continued, and if all the cache servers have been completed, the processing is terminated.
[0166]
FIG. 42 is a flowchart showing the processing of the Dijkstra calculation means 210. The Dijkstra calculation unit 210 sets the edge information of the cache server (information of the edge node to which the cache server is connected) at the start of the Dijkstra calculation (S21001).
[0167]
Next, the Dijkstra calculation unit 210 reads out the link information of the processing target edge node from the network statistical information DB 303 (S21001). Next, the Dijkstra calculation unit 210 determines whether there is a vacancy for the requested bandwidth (S21003). If there is no free space (S21003; NO), the Dijkstra calculating means 210 skips the process, and if there is a free space (S21003; YES), determines whether or not the connection destination of the link is other than the non-passing edge node. (S21004).
[0168]
At this time, the Dijkstra calculation means 210 skips the processing if the connection destination of the link is not a non-passing edge node (S21004; NO), and if the link is not a non-passing edge node (S21004; YES), the link is deactivated. It is determined whether or not the link is other than the passing link (S21005).
[0169]
At this time, the Dijkstra calculating means 210 skips the process if the link destination is not a non-passing link (S21005; NO), and exceeds the maximum hop number if the link is not a non-passing link (S21005; YES). It is determined whether or not it has not been performed (S21006). At this time, the Dijkstra calculation means 210 skips the processing if the number of hops exceeds the maximum hop number (S21006; NO), and calculates the link cost if the number does not exceed the maximum hop number (S21006; YES) (S21007).
[0170]
Next, the Dijkstra calculation unit 210 determines whether the link cost calculated in S21007 is smaller than the cost value processed last time (S21008). The Dijkstra calculation means 210 skips the processing if the link cost is higher than the previous one (S21008; NO), and replaces the link cost value if the link cost is lower (S21008; YES) (S21009).
[0171]
Next, the Dijkstra calculation means 210 determines whether or not processing has been performed for all links (S21010). If there is any remaining (S21010; NO), the process is continued, and if all the processes have been completed (S21010; YES), the corresponding edge node is determined as a route (S21011).
[0172]
Next, the Dijkstra calculation unit 210 determines whether or not processing has been performed on all edge nodes (S21012). At this time, the Dijkstra calculation means 210 continues the processing if there is any remaining (S21012; NO), and if all the processing has been completed (S21012; YES), obtains the route information (S21013) and ends the processing.
[0173]
As described above, the Dijkstra calculation unit 210 searches for the optimal route connecting the cache server edge and the user edge by using the Dijkstra method. The path information is passed to 206 along with the determination result of OK. On the other hand, if the optimum route cannot be found, the result of the NG determination is passed to the route search means 206.
[0174]
FIG. 43 is a flowchart showing the processing of the QoS path management means (QoS path management unit) 207. The QoS path management unit 207 converts the user edge IP address and the server edge IP address passed by the QoS path management parameter 48 for the network including the network device A, the network device B, the network device C, and the network device D. A request for quality assurance in End-To-End is sent to the network control means 208.
[0175]
At this time, the QoS path management means 207 gives the network control parameters 208 (FIG. 44) to the network control means 208 as input parameters. The network control parameters 52 include a user IP address, a user edge IP address, a reserved band, and a content ID. Then, a result indicating whether or not the network control has been normally completed is returned as a network control response parameter 53 (FIG. 45). The network control response parameter 53 includes a response result (OK / NG) of the bandwidth registration request.
[0176]
In FIG. 43, the QoS path management unit 207 first refers to the use frequency level in the QoS path management parameter 48, and stores the QoS path aggregation method information 51 (prepared for each use frequency level in the network / resource management DB 302). From FIG. 46), QoS path aggregation method information corresponding to the use frequency level in the parameter 48 is extracted (S20701). As shown in FIG. 46, the QoS path aggregation method information 51 can include a usage frequency level and a QoS path setting method flag as a setting method flag related to the QoS path setting. For example, if the flag is off, an “on-demand” QoS path “without flow aggregation” is set, and if the flag is on, an “on-demand” QoS path with “flow aggregation” is set.
[0177]
Next, the QoS path management unit 207 determines whether or not to perform flow aggregation according to the extracted QoS path aggregation method information (level and flag values) (S20702). At this time, when performing flow aggregation (S20702; YES), the QoS path management unit 207 sets an aggregation flow flag (S20703), and advances the processing to S20705. If flow aggregation is not to be performed (S20702; NO), the process proceeds to S20705.
[0178]
Next, the QoS path management unit 207 determines whether or not the aggregation flow flag is set (S20705). At this time, if the aggregation flow flag is set (S20705; YES), the QoS path management unit 207 sets the QoS path with “on-demand flow aggregation” (for example, level 2 in FIG. 4). For this purpose, the process is passed to the network control means 208 (S20706).
[0179]
On the other hand, if the aggregation flow flag is not set (S20705; NO), the QoS path management unit 207 sets the QoS path without on-demand type flow aggregation (for example, level 1 in FIG. 4). Then, the process is passed to the network control unit 208 (S20707).
[0180]
As described above, the QoS path management unit 207 changes the QoS path collection method information, thereby enabling flow aggregation and setting a QoS path on demand. Then, when the QoS path setting method is determined, the QoS path management unit 207 issues a request to perform quality assurance by End-To-End with the user edge IP address and the server edge IP address passed in the QoS path management parameter 48. Request to control means 208.
[0181]
FIG. 48 is a flowchart showing the processing of the network control means (network control unit) 208. In response to a request from the QoS path management unit 207, the network control unit 208 sends the user edge passed as the network control parameter 52 to the network including the network device A, the network device B, the network device C, and the network device D. Quality is guaranteed between the IP address and the server edge IP address by End-To-End (S20801).
[0182]
That is, the network control unit 208 sets a QoS path corresponding to a use frequency level that can accommodate a GS flow in which at least the band specified by the use request (reserved band specified in the network control parameter 52) is secured, It is set by controlling the edge node, the server edge node, and the passing edge node between them. Then, network control parameters 53 including whether or not the QoS path can be set (response result (OK / NG) of the bandwidth registration request) are passed to the QoS path management unit 207.
[0183]
At this time, the QoS path management means 207 gives a response result (OK / NG) according to the response result from the network control means 208 to the QoS path setting means 204. The QoS path setting unit 204 returns a response result (OK / NG) corresponding to the response result from the QoS path management unit 207 to the usage status determining unit 203 (S20408).
[0184]
The flow (traffic flow) relating to the service from the user can be released (released) by the release processing by the QoS path setting means 204, triggered by the release request from the user or the end of the service provision. At this time, when the flow flowing in the QoS path becomes 0, the QoS path can be released.
[0185]
On the other hand, in order to perform the quality control as shown in FIG. 4, the flow aggregation marking means (flow aggregation marking unit) 209 performs the following processing. FIG. 49 is a flowchart showing the processing of the flow aggregation marking means 209.
[0186]
The flow aggregation marking means 209 is activated by the reception control means 205 when the bandwidth related to the user's use request exceeds the reserved bandwidth (the bandwidth that can be aggregated by the already set QoS path) (S20502). The flow aggregation marking unit 209 sets the reserved bandwidth over flag (FIG. 40) of the QoS path setting information 50 stored in the network / resource management DB 302 to ON (S20901), and ends the processing.
[0187]
For a QoS path for which the reserved band over flag is set to ON, the QoS path is not released even if the number of active users becomes zero. As a result, a flow for the next use request can be allocated to the QoS path. It is the QoS path period releasing means 212 that releases such a QoS path.
[0188]
FIG. 50 is a flowchart showing the processing of the QoS path cycle releasing means (QoS path cycle releasing unit) 212. The QoS path period releasing means 212 starts periodically, refers to the QoS path setting information 50 stored in the network resource management DB 302 in order from the top, sets the reserved band over flag to ON, and sets the number of in-use users to 0. It is determined whether or not there is a QoS path of (S21201). Then, if there is a corresponding QoS path, the QoS path period releasing unit 212 releases the QoS path (S21202).
[0189]
In this way, the QoS path for which the reserved band over flag is set to ON remains effective until released by the QoS path period releasing unit 212 even if the number of in-use users becomes zero. Then, when the flow relating to the next use request is set in the QoS path, the number of users in use exceeds 0, so the QoS path period releasing unit 212 does not release the QoS path.
[0190]
FIG. 51 is a flowchart showing the processing of the QoS path aggregation flow management unit (QoS path aggregation flow management unit) 213. The QoS path aggregation flow management unit 213 subtracts the reserved band (used band) from the reserved band in the QoS path setting information 50 stored in the network resource management DB 302 (S21301).
[0191]
Next, the QoS path aggregation flow management unit 213 determines whether or not the subtracted value is smaller than a set threshold value (prepared in advance) (S21302). At this time, if the subtraction value is not smaller (larger) than the threshold value (S21302; NO), the QoS path aggregated flow management unit 213 sets a smaller reserved bandwidth when securing the aggregated flow (S21303).
[0192]
Next, the QoS path aggregation flow management unit 213 determines whether all the QoS path setting information 50 has been checked (S21304). At this time, if the check is not completed (S21304; NO), the process is continued, and if it is completed (S21304; YES), the process is ended.
[0193]
In this manner, when the available bandwidth of the QoS path is larger than the threshold, the QoS path aggregated flow management unit 213 reduces the set value of the reserved bandwidth of the QoS path of the aggregated flow to be secured next. Thereby, the band reserved in advance can be reduced.
[0194]
FIG. 52 is a flowchart showing the processing of the load distribution calculation means (load distribution calculation unit) 214. In the calculation by the load distribution calculation unit 214, in the conventional method, the calculation of the free bandwidth of the link is performed using only the actual traffic in which the GS flow and the BES flow are mixed.
[0195]
On the other hand, the load balancing calculation unit 214 in the embodiment uses a QoS path (marked by the flow aggregation marking unit 209) indicating that the usage frequency level of the QoS path setting information 50 managed by the network / resource management DB 302 is high. For the “QoS path with flow aggregation” on which the processing has been performed, the calculation of the available bandwidth of the link is performed by taking into account the amount obtained by subtracting the used bandwidth from the reserved bandwidth (the available bandwidth in the bandwidth for quality assurance) as predicted traffic. Performed (S21001).
[0196]
In this way, the load distribution calculation unit 214 calculates the free bandwidth of each link in a plurality of routes in the network, and determines the traffic distribution ratio for each route based on the free bandwidth. Then, the load distribution control unit 216 executes the load distribution processing (allocation of traffic to routes) so that the traffic flows at the determined distribution ratio (for example, so that the distribution ratio is within a desired range (for example, averaging)). )I do. It is preferable that the processing of the load distribution calculation unit 214 (determination of the distribution ratio) be performed, for example, periodically. Further, for example, non-guaranteed traffic may be generated, and the determination may be made when deciding which route to allocate this non-guaranteed traffic to.
[0197]
According to the load distribution calculation means 214, the distribution ratio is determined so that traffic flows through a plurality of routes at a desired distribution ratio by load distribution. At this time, in the definition of the free bandwidth when determining the distribution ratio, the free bandwidth of the QoS path expected to be used in the near future is considered. Then, the available bandwidth of the link constituting each route is determined, and the distribution ratio is determined based on the determined available bandwidth of the link.
[0198]
As a result, for example, after a non-guaranteed traffic flows on a link of a certain path (a QoS path capable of flow aggregation has already been set) based on the calculation of the available bandwidth as in the related art, the QoS path is guaranteed. By setting the traffic (flow), it is possible to prevent the band used by the link from increasing and the traffic distribution ratio from being biased. Also, by setting guaranteed traffic on the QoS path, it is possible to suppress occurrence of congestion in non-guaranteed traffic.
[0199]
Further, the load distribution calculation unit 214 may be configured to extract a QoS path that is likely to use an available bandwidth from the QoS paths in the link, and to consider the available bandwidth only for the extracted QoS path. it can. For example, the load balancing calculation unit 214 extracts a QoS path (marked QoS path) capable of flow aggregation set in the link when obtaining an available bandwidth of the link of the route, and Find the current usage frequency level. At this time, when the usage frequency level is low (for example, level 1 in FIGS. 3 and 4), the load balancing calculation unit 214 determines that the state of the QoS path corresponds to the case C shown in FIGS. And does not take into account the subtraction of the available bandwidth of the QoS path from the link bandwidth. On the other hand, when the current usage frequency level is high (for example, level 2 in FIGS. 3 and 4), the load balancing calculation unit 214 determines that the state of the QoS path corresponds to the case B shown in FIGS. Therefore, the available bandwidth of the QoS path is taken into account in the subtraction from the link bandwidth.
[0200]
<Modified example>
The embodiment described above can be modified as follows. The operations of the QoS path management means 207, the network control means 208, the flow aggregation marking means 209, and the QoS path cycle release means 212 in the above embodiment are performed assuming the quality assurance control method shown in FIG. On the other hand, the quality assurance control method shown in FIG. 3 is realized as follows.
[0201]
The operations at the usage frequency levels 1 and 2 are as described above. On the other hand, the quality assurance control at the use frequency levels 3 and 4 is realized by the following operation by the QoS path cycle management unit 211.
[0202]
FIG. 47 is a flowchart showing the processing of the QoS path cycle management means (QoS path cycle management unit) 211. The QoS path cycle management means 211 starts periodically (periodically), and obtains the current usage frequency level by referring to the service usage management DB 301 (S21100). The use frequency level can be obtained by obtaining a use frequency value by the above-described method and obtaining a use frequency level corresponding to the use frequency value.
[0203]
Next, the QoS path cycle management unit 211 determines whether the current level is 3 or 4 (S21101). At this time, if the level is level 4, the process is passed to the network control unit 208 in order to perform QoS path setting by “reservation type flow aggregation” (S21102). On the other hand, if the level is level 3, the QoS path cycle management unit 211 passes the process to the network control unit 208 to perform the QoS path setting without “reservation type flow aggregation” (S21103). Thereafter, the network control unit 208 sets the reservation type QoS path according to the network control parameters (data such as the bandwidth of the setting target QoS path) for setting the reservation type QoS path. As a result, when the usage frequency level becomes 3 or 4, the reservation type QoS path is already set (there is a vacant band). Therefore, when there is a use request from the user at level 3 or 4, the determination of YES is always made in S20502 (FIG. 39) of the reception control unit 205, and the band in the empty state (reserved QoS) Path), the flow related to the use request is allocated.
[0204]
The operation of the QoS path cycle management means 211 can be configured to be turned on / off according to the setting. Also, the flow aggregation marking means 209 and the QoS path cycle releasing means 212 may be turned off in response to the turning on of the QoS path cycle managing means 211. Further, the QoS path cycle management unit 211, the flow aggregation marking unit 209, and the QoS path cycle release unit 212 may be configured to operate.
[0205]
Further, as another embodiment, a configuration as shown in FIGS. 53 and 54 can be applied. In FIG. 53, the QoS path management unit 207 determines whether or not the usage frequency level is a level at which a reservation is made by determining whether or not the usage frequency level is higher than a predetermined level (for example, level 3 ′, level 4 ′). It is determined whether or not it is (S20704). If it is the level at which the reservation is made, the QoS path management means 207 activates the QoS cycle management means 211 (S20708), and the QoS path of reservation type flow aggregation as shown in FIG. (QoS path) or a QoS path without reservation type flow aggregation (QoS path according to level 3 '). Further, the QoS path management unit 207 converts the on-demand type QoS path for accommodating the flow according to the use request to the use frequency level (3 ′ (no flow aggregation), level 4 ′ (with flow aggregation)). Therefore set. In the embodiment shown in FIG. 53, when the use frequency is level 1 or 2, the same operation as the operation shown in FIG. 43 is performed.
[0206]
In addition, the QoS path aggregation flow management unit 213 can be configured to perform processing for increasing the bandwidth of the reserved bandwidth as follows. That is, the network resource management DB 302 is configured to store a predetermined number of times (marking times) when marking is performed on the QoS path by the flow aggregation marking unit 209. Then, the QoS path aggregation flow management means 213 is periodically activated, and obtains the number of markings per hour (marking frequency) using the marking time stored in the network resource management DB 302. The marking frequency can be obtained by the same method as the usage frequency. Then, when the marking frequency exceeds a predetermined threshold value prepared in advance, the QoS path aggregated flow management unit 213 increases the value of the reserved bandwidth for securing the aggregated flow in the QoS path setting information 50. Such a process can be performed together with or instead of the process shown in FIG.
[0207]
[Effects of the embodiment]
According to the embodiment, when providing a broadband information service typified by VOD or live broadcasting that requires quality control in a network, or a two-way voice communication service typified by VoIP, a content provider or a network The operator does not need to set very complicated quality control for the network device (router). Therefore, the content provider and the network operator can easily operate and manage the service without having high skills. As a result, operation costs including labor costs and the like can be significantly reduced.
[0208]
When the quality assurance request reception method (admission control method) is used, the quality control in the network is set for each end-to-end flow for only the services that are not frequently used. As a result, control over the network device (router) is reduced. Further, it is possible to improve the allowable number of services due to the upper limit that can be set for each end-to-end flow in the network device (router).
[0209]
On the other hand, using a quality assurance request reception method (admission control method) using a flow aggregation method, cooperates with a broadband information service represented by VOD or live broadcasting, and a two-way voice communication service represented by VoIP, etc. Even in the case of performing effective quality assurance in a network, the flow aggregation method can be applied only to those having a high frequency of service use.
[0210]
Further, when a request exceeding the upper limit of the network resources secured by the flow aggregation is received, it is searched for a free path in another path (route), and an alternative path (route) can be used. Resources can be used effectively.
[0211]
As a secondary effect, carriers, ISPs, IDCs, and the like can easily realize quality assurance in a broadband information service represented by VOD or live broadcasting and a two-way voice communication service represented by VoIP. Services that can collect service usage fees can be provided, enabling new business development.
[0212]
In addition, non-guaranteed traffic, which is a traffic that must be considered by carriers, ISPs, and IDCs that provide broadband information services represented by VOD and live broadcasting, and two-way voice communication services represented by VoIP, etc. , As much as possible, and effective dynamic load distribution using multipath can be realized.
[0213]
[Others]
The embodiment described above discloses the following invention. Further, the following inventions can be appropriately combined as needed.
(Appendix 1) Installed in a network composed of network devices with a quality assurance function,
Use request frequency acquisition means for obtaining a use request frequency as a frequency of a use request for a network service including a quality assurance request from a user;
Control means for performing quality assurance network control based on the use request frequency,
Including quality assurance control equipment. (1)
(Supplementary Note 2) The control means performs a quality assurance network control for setting a network resource for quality assurance to which a traffic flow for a certain use request can be allocated together with another traffic flow, and a use request generated thereafter. Traffic flows to the network resources for quality assurance
The quality assurance control device according to attachment 1. (2)
(Supplementary Note 3) The control unit performs quality assurance network control for securing a network resource for quality assurance for allocating a traffic flow for one use request.
The quality assurance control device according to attachment 1 or 2.
(Supplementary Note 4) The control means performs a quality assurance network control for reserving (pre-securing) a network resource for quality assurance to which a plurality of traffic flows can be allocated, and controlling a traffic flow for a use request generated thereafter. Allocate to pre-allocated network resources for quality assurance
The quality assurance control device according to any one of supplementary notes 1 to 3. (3)
(Supplementary Note 5) The control unit performs quality assurance network control for reserving (preliminarily securing) a network resource for quality assurance to which a traffic flow for one use request can be allocated, and for traffic for a use request generated thereafter. Allocate flows to network resources reserved for quality assurance in advance
The quality assurance control device according to any one of supplementary notes 1 to 4.
(Supplementary Note 6) The use request frequency is classified into a plurality of levels,
The control unit performs the quality assurance network control according to a level corresponding to a current use request frequency.
The quality assurance control device according to any one of supplementary notes 1 to 5.
(Supplementary Note 7) If the traffic flow for the newly accepted use request cannot be allocated to the already set network resource for quality assurance, the control unit assigns the traffic flow to the use request. Secure new assignable network resources for quality assurance
The quality assurance control device according to attachment 2 or 4. (4)
(Supplementary Note 8) Even when there is no traffic flow allocated to the quality assurance network resource, the control means maintains the quality assurance network resource for a certain period of time, during which a new use request is issued. When it occurs, the traffic flow for this use request is allocated to the network resource for quality assurance.
The quality assurance control device according to attachment 7. (5)
(Supplementary Note 9) The control unit monitors an available bandwidth of the network resource for quality assurance, and when the available bandwidth exceeds a predetermined threshold, a network resource for quality assurance newly secured. Decrease the bandwidth setting
The quality assurance control device according to attachment 7 or 8. (6)
(Supplementary Note 10) When the frequency of the new setting of the network resource for quality assurance exceeds a predetermined threshold, the control unit changes the set value of the bandwidth of the newly secured network resource for quality assurance. Enlarge
The quality assurance control device according to any one of supplementary notes 7 to 9.
(Supplementary Note 11) A load distribution device for dynamically distributing traffic flowing through a network to a plurality of routes,
Free bandwidth calculation means for calculating the free bandwidth of the link of each route;
Means for determining a distribution ratio of traffic to the plurality of routes based on an available bandwidth of a link of each route,
When a band for quality assurance is secured in the link, the free band calculation means subtracts the free band in the band for quality assurance and the actually used band of the link from the link band, and A load balancer that seeks free bandwidth. (7)
(Supplementary Note 12) The vacant band calculation means subtracts the vacant band in the band for quality assurance from the link band only when it is determined that the vacant band in the band for quality assurance is likely to be used.
The load distribution device according to attachment 11. (8)
(Supplementary Note 13) The vacant bandwidth of the bandwidth for quality assurance is monitored, and if the vacant bandwidth exceeds a predetermined threshold, the set value of the bandwidth for quality assurance newly set for the link is reduced. Further includes means
13. The load distribution device according to supplementary note 11 or 12.
[0214]
In such a configuration, when a network control of quality assurance that enables aggregation of a plurality of traffic flows is performed, an unused band, which is a difference between a reserved band and a reserved band, is periodically reserved. It is possible to monitor whether the traffic flow is not left, and when the traffic flow is excessively secured, it is possible to reduce a reserved bandwidth that can aggregate a plurality of traffic flows.
(Supplementary Note 14) If the frequency of newly securing a band for quality assurance for the link exceeds a predetermined threshold, a means for increasing the set value of the band for quality assurance newly set for the link is further provided. Including
14. The load distribution device according to any one of supplementary notes 11 to 13.
[0215]
In such a configuration, when the allowable capacity of the service use request that can ensure the quality assurance, for example, frequently exceeds the secured bandwidth, the reserved bandwidth that can aggregate a plurality of traffic flows is increased. be able to.
(Supplementary Note 15) A quality assurance control device provided in a network including network devices having a quality assurance function
Obtain the usage request frequency as the frequency of usage requests for network services including quality assurance requests from users,
Perform quality assurance network control according to the usage request frequency
Quality assurance control method. (9)
(Supplementary Note 16) Based on the use request frequency, a quality assurance network control for securing a network resource for quality assurance that can allocate a traffic flow to a certain use request and another traffic flow is performed. A traffic flow for the requested use request to the network resource for quality assurance
The quality assurance control method according to supplementary note 15.
(Supplementary Note 17) Based on the use request frequency, quality assurance network control for securing a network resource for quality assurance for allocating a traffic flow for one use request is performed.
The quality assurance control method according to attachment 15 or 16.
(Supplementary Note 18) Based on the use request frequency, a quality assurance network control for reserving (preliminarily securing) a network resource for quality assurance to which a plurality of traffic flows can be allocated, Allocate flows to network resources reserved for quality assurance in advance
18. The quality assurance control method according to any one of supplementary notes 15 to 17.
(Supplementary Note 19) Based on the use request frequency, quality assurance network control for reserving (preliminarily securing) a network resource for quality assurance for allocating a traffic flow for one use request is performed. Allocate the traffic flow for the request to pre-reserved network resources for quality assurance
19. The quality assurance control method according to any one of supplementary notes 15 to 18.
(Supplementary Note 20) The use request frequency is classified into a plurality of levels,
The control unit performs the quality assurance network control according to a level corresponding to a current use request frequency.
20. The quality assurance control method according to any one of supplementary notes 15 to 19.
(Supplementary Note 21) If the traffic flow for the newly accepted use request cannot be assigned to the already set network resource for quality assurance, a new traffic flow for assigning the traffic flow to the use request can be assigned. Secure network resources for quality assurance
The quality assurance control method according to attachment 16 or 18.
(Supplementary Note 22) Even when there is no longer a traffic flow allocated to the quality assurance network resource, the control means maintains the quality assurance network resource for a certain period of time, during which a new use request is issued. When it occurs, the traffic flow for this use request is allocated to the network resource for quality assurance.
The quality assurance control method according to attachment 21.
(Supplementary Note 23) The control unit monitors an available bandwidth of the network resource for quality assurance, and when the available bandwidth exceeds a predetermined threshold, a newly set network resource for quality assurance. Decrease the bandwidth setting
The quality assurance control method according to attachment 21 or 22.
(Supplementary Note 24) When the frequency of setting a new network resource for quality assurance exceeds a predetermined threshold value, the control unit may change the set value of the bandwidth of the newly set network resource for quality assurance. Enlarge
24. The quality assurance control method according to any one of supplementary notes 21 to 23.
(Supplementary Note 25) A load distribution device that dynamically distributes traffic flowing through a network to a plurality of routes is provided by:
When calculating the available bandwidth of the link of each route, if a bandwidth for quality assurance is secured for the target link, the available bandwidth in the bandwidth for quality assurance and the actually used bandwidth of the link are used. Subtract the link bandwidth to determine the free bandwidth of the link,
Determining a traffic distribution ratio for the plurality of routes based on an available bandwidth of a link of each route;
And a load balancing method. (10)
(Supplementary Note 26) Only when it is determined that a vacant band in the band for quality assurance is likely to be used, this vacant band is subtracted from the link band.
The load distribution method according to attachment 25.
(Supplementary Note 27) The vacant band in the band for quality assurance is monitored, and when the vacant band exceeds a predetermined threshold, the bandwidth of the network resource for quality assurance newly set to the link is monitored. Decrease the set value
27. The load distribution method according to supplementary note 25 or 26, further comprising:
(Supplementary Note 28) If the frequency of setting a new bandwidth for quality assurance for the link exceeds a predetermined threshold, the set value of the bandwidth of the network resource for quality assurance newly set for the link is increased. Do
28. The load balancing method according to any one of supplementary notes 25 to 27, further comprising:
(Supplementary Note 29) A recording medium recording a program for causing a computer to function as a quality assurance control device provided in a network including network devices having a quality assurance function,
Obtaining a usage request frequency as a frequency of a usage request for a network service including a quality assurance request from a user;
Performing quality assurance network control according to the use request frequency;
Recording medium that stores a program for causing a computer to execute the program.
(Supplementary Note 30) A recording medium recording a program that causes a computer to function as a load distribution device that dynamically distributes traffic flowing through a network to a plurality of routes,
When calculating the available bandwidth of the link of each route, if a bandwidth for quality assurance is secured for the target link, the available bandwidth in the bandwidth for quality assurance and the actually used bandwidth of the link are used. Subtracting from the link bandwidth to determine the free bandwidth of the link;
Determining a traffic distribution ratio for the plurality of routes based on an available bandwidth of a link of each route;
Recording medium that stores a program for causing a computer to execute the program.
(Supplementary Note 31) A program for causing a computer to function as a quality assurance control device provided in a network including network devices having a quality assurance function,
Obtaining a usage request frequency as a frequency of a usage request for a network service including a quality assurance request from a user;
Performing quality assurance network control according to the use request frequency;
A program that causes a computer to execute.
(Supplementary Note 32) A program that causes a computer to function as a load distribution device that dynamically distributes traffic flowing through a network to a plurality of routes,
When calculating the available bandwidth of the link of each route, if a bandwidth for quality assurance is secured for the target link, the available bandwidth in the bandwidth for quality assurance and the actually used bandwidth of the link are used. Subtracting from the link bandwidth to determine the free bandwidth of the link;
Determining a traffic distribution ratio for the plurality of routes based on an available bandwidth of a link of each route;
A program that causes a computer to execute.
[0216]
【The invention's effect】
ADVANTAGE OF THE INVENTION According to this invention, the quality assurance control technique which can reduce the burden of a network service provider can be provided.
[0219]
Further, according to the present invention, it is possible to provide a quality assurance control technique capable of improving prompt response to a use request.
[0218]
Further, according to the present invention, it is possible to provide a load distribution technique capable of determining a distribution ratio for a plurality of routes in consideration of quality assurance traffic.
[Brief description of the drawings]
FIG. 1 is a conventional principle diagram related to the present invention.
FIG. 2 is a principle diagram of the present invention.
FIG. 3 is a diagram illustrating a provisioning method according to a service use situation;
FIG. 4 is a diagram illustrating a flow reuse type provisioning method according to a service use situation.
FIG. 5 is an explanatory diagram showing a relationship between a QoS path (GS flow) and non-guaranteed traffic (BES flow).
FIG. 6 is an explanatory diagram of a level transition of a QoS path.
FIG. 7 is a diagram illustrating a link having a plurality of aggregated paths;
FIG. 8 is a system configuration diagram for explaining an embodiment of the present invention.
FIG. 9 is a login screen according to the embodiment of the present invention.
FIG. 10 is a diagram showing a display example of a service content list screen according to the embodiment of the present invention.
FIG. 11 is a content reproduction screen according to the embodiment of the present invention.
FIG. 12 is a flowchart of a quality assurance control process from a user in the embodiment of the present invention.
FIG. 13 is an explanatory diagram of a login message.
FIG. 14 is an explanatory diagram of user information.
FIG. 15 is an explanatory diagram of content information.
FIG. 16 is an explanatory diagram of a content selection message.
FIG. 17 is an explanatory diagram of a quality assurance request message.
FIG. 18 is an explanatory diagram of a quality assurance response message.
FIG. 19 is an explanatory diagram of a content selection message.
FIG. 20 is an explanatory diagram of a processing flow of a quality assurance request receiving means 200 of the quality assurance server.
FIG. 21 is an explanatory diagram of quality assurance parameters.
FIG. 22 is an explanatory diagram of a processing flow of quality assurance request analysis means;
FIG. 23 is an explanatory diagram of a usage status accumulation parameter;
FIG. 24 is an explanatory diagram of QoS path setting release parameters.
FIG. 25 is an explanatory diagram of a processing flow of a usage status accumulating unit;
FIG. 26 is an explanatory diagram of a service usage status management DB;
FIG. 27 is an explanatory diagram of a usage status determination request parameter;
FIG. 28 is an explanatory diagram of a processing flow of a usage status determination unit;
FIG. 29 is an explanatory diagram of a processing flow of a QoS path setting unit.
FIG. 30 is an explanatory diagram of reception control parameters.
FIG. 31 is an explanatory diagram of a reception control response parameter.
FIG. 32 is an explanatory diagram of a network resource management DB.
FIG. 33 is an explanatory diagram of cache server information stored in a network / resource management DB;
FIG. 34 is an explanatory diagram of edge information stored in a network resource management DB;
FIG. 35 is an explanatory diagram of a route search parameter.
FIG. 36 is an explanatory diagram of a route search response parameter.
FIG. 37 is an explanatory diagram of QoS path management parameters.
FIG. 38 is an explanatory diagram of a QoS path management response parameter.
FIG. 39 is an explanatory diagram of a processing flow of a reception control unit;
FIG. 40 is an explanatory diagram of QoS path setting information stored in a network resource management DB.
FIG. 41 is an explanatory diagram of a processing flow of a route search unit;
FIG. 42 is an explanatory diagram of a processing flow of Dijkstra calculation means;
FIG. 43 is an explanatory diagram of a processing flow of a QoS path management unit;
FIG. 44 is an explanatory diagram of network control parameters.
FIG. 45 is an explanatory diagram of a network control response parameter.
FIG. 46 is an explanatory diagram of QoS path aggregation method information stored in the network resource management DB.
FIG. 47 is an explanatory diagram of a processing flow of QoS path cycle management means;
FIG. 48 is an explanatory diagram of a processing flow of a network control unit;
FIG. 49 is an explanatory diagram of the processing flow of the flow aggregation marking means.
FIG. 50 is an explanatory diagram of a processing flow of QoS path cycle releasing means;
FIG. 51 is an explanatory diagram of a processing flow of QoS path aggregation flow management means;
FIG. 52 is an explanatory diagram of a processing flow of a load distribution calculation unit;
FIG. 53 is an explanatory diagram of another processing flow of the QoS path management means;
FIG. 54 is an explanatory diagram of another processing flow of the QoS path cycle management unit;
[Brief description of reference numerals]
A to D, 101 to 104 Network equipment
100 networks
105 clients
106 servers
106A, 106B VOD server
107 Quality Assurance Server
108 Service reception server
200 Quality Assurance Request Acceptance Method
201 Quality assurance requirement analysis means
202 Usage status storage means
203 Usage status determination means
204 QoS path setting means
205 reception control means
206 Route search means
207 QoS path management means
208 Network control means
209 Flow aggregation marking means
210 Dijkstra calculation means
211 QoS path cycle management means
212 QoS path cycle release means
213 QoS path aggregation flow management means
214 load balancing calculation means
215 Network traffic measurement means
216 Load distribution control means
301 Service Usage Status Management DB
302 Network Resource Management DB
303 Network Statistics DB

Claims (10)

Installed in a network consisting of network devices with quality assurance functions,
Use request frequency acquisition means for obtaining a use request frequency as a frequency of a use request for a network service including a quality assurance request from a user;
Control means for performing quality assurance network control based on the use request frequency,
Including quality assurance control equipment. The control means performs quality assurance network control for securing a network resource for quality assurance that can allocate a traffic flow to another use flow together with a traffic flow to a certain use request. 2. The quality assurance control device according to claim 1, wherein the network resource is allocated to the network resource for quality assurance. The control means performs a quality assurance network control for reserving a network resource for quality assurance to which a plurality of traffic flows can be allocated, and a network for quality assurance in which a traffic flow for a use request generated thereafter is reserved. 3. The quality assurance control device according to claim 1, wherein the quality assurance control device is assigned to a resource. If the control means cannot allocate the traffic flow for the newly accepted use request to the already set network resource for quality assurance, 4. The quality assurance control device according to claim 2, wherein a network resource for quality assurance is secured. The control means maintains the network resource for quality assurance for a certain period of time even when there is no traffic flow allocated to the network resource for quality assurance, and when a new use request occurs during that time, 5. The quality assurance control device according to claim 4, wherein a traffic flow corresponding to the use request is allocated to the quality assurance network resource. The control means monitors an available bandwidth of the quality assurance network resource, and when the available bandwidth exceeds a predetermined threshold, sets a newly set quality assurance network resource bandwidth. The quality assurance control device according to claim 4 or 5, wherein the value is reduced. A load balancing device for dynamically distributing traffic flowing through a network to a plurality of routes,
Free bandwidth calculation means for calculating the free bandwidth of the link of each route;
Means for determining a distribution ratio of traffic to the plurality of routes based on an available bandwidth of a link of each route,
When a band for quality assurance is secured in the link, the free band calculation means subtracts the free band in the band for quality assurance and the actually used band of the link from the link band, and A load balancer that seeks free bandwidth. The said free band calculation means subtracts the free band in the band for quality assurance from the link band only when it is judged that there is a high possibility that the free band in the band for quality assurance is used. Load balancer. A quality assurance control device provided in a network composed of network devices with a quality assurance function,
Obtain the usage request frequency as the frequency of usage requests for network services including quality assurance requests from users,
A quality assurance control method including performing quality assurance network control according to the use request frequency. A load balancer that dynamically distributes traffic flowing through the network to multiple routes,
When calculating the available bandwidth of the link of each route, if a bandwidth for quality assurance is secured for the target link, the available bandwidth in the bandwidth for quality assurance and the actually used bandwidth of the link are used. Subtract the link bandwidth to determine the free bandwidth of the link,
A load distribution method, comprising: determining a traffic distribution ratio for the plurality of routes based on an available bandwidth of a link of each route.

Images