WO2007051349A1

WO2007051349A1 - A multi-services internet protocol bearer system having quality of service assurance and implementing method thereof

Info

Publication number: WO2007051349A1
Application number: PCT/CN2005/001813
Authority: WO
Inventors: Chen Wu; Hongbiao Li; Ke Du; Xianmin Meng
Original assignee: Zte Corporation
Priority date: 2005-11-01
Filing date: 2005-11-01
Publication date: 2007-05-10
Also published as: CN101228764B; CN101228764A

Abstract

A multi-services IP bearer system having QoS assurance and the implementing method thereof, it includes at least : a switching function unit, which supports the resource scheduling and isolation, and support to match the specific service with the specific network resource and forward the specific service using the specific network resource; and resource control function unit, which controls and manages the resource scheduling of switching function unit, and downward-transfer the resource scheduling and the allocation rule as well as the mapping relation of the resource and the service to switching function unit. The logic networks loaded service by the bearer system are separated each other, QoS of various services is not affected from other services, and it can receive QoS feedback of the service network, schedule the resource in the IP bearer system, and improve the QoS assurance for services.

Description

Multi-service internet protocol bearer system with service quality guarantee and implementation method thereof

The present invention relates to a network system capable of carrying multiple services by using an Internet Protocol (IP) technology, in particular, a multi-service IP bearer system with quality of service (QoS) guarantee and implementation. method. Background technique

The inherent technical and economic characteristics of IP technology have created the brilliance of the Internet, prompting the telecommunications industry to try to increase the number of telecom services, such as softswitch, Internet Protocol Television (IPTV), The 3rd Generation (hereinafter referred to as 3G:) and the Next Generation Network (NGN) are carried on the IP network. However, the connectionless operation mode of IP technology, best-effort transmission, bandwidth statistical multiplexing, and hop-by-hop congestion packet discarding processing make the IP network QoS guarantee capability have defects. Existing QoS guarantee schemes, such as Integrated Services (IntServ) model, Differentiated Service (DiffServ) model, and traffic engineering, or cannot be deployed on the entire network for scalability reasons, or because management granularity Roughness does not solve the end-to-end QoS guarantee problem.

In recent years, more research has been done on resource admission control subsystems (Resource and Admission).

Control SubsystemX (hereinafter referred to as RACS) technology proposes a layer-to-layer interaction mode between the service layer and the bearer layer, which increases the complexity of carrying multiple services on the IP bearer network, and when there is burst traffic in the IP bearer network. , business interference with each other, RACS technology QoS assurance capabilities are also worthy of further research and validation of the invention

In view of the above, it is an object of the present invention to provide a multi-service IP bearer system with QoS guarantee and an implementation method thereof. The IP bearer system and the upper layer service network have a clear interface, and the IP bearer system ensures the QoS of the service network within the rated service traffic range, and each service network controls the service within the allowed service traffic range to enter the IP bearer. system. The IP bearer system can accept QoS feedback from the service network, schedule resources in the IP bearer system, and improve QoS guarantee for services.

In the multi-service IP bearer system with QoS guarantee proposed by the present invention, the logical networks of the carried services are isolated from each other, and the QoS of various services is not affected by other services.

To achieve the above objective, a multi-service IP bearer system with QoS guarantee of the present invention includes at least:

a switching functional unit that supports resource scheduling and isolation, and supports matching specific services with specific network resources and forwarding the particular service using the particular network resource;

a resource control function unit, which controls and manages the resource scheduling of the switching function unit, and sends the resource scheduling and distribution rules and the mapping relationship between the resources and the service to the switching function unit;

The foregoing switching function unit includes a router and a switch. The resource control function unit includes a service flow requirement table and a network resource allocation table, where the service flow requirement table is an end-to-end service flow requirement table of the service network to the IP bearer system. The foregoing service network records the service traffic and flow direction carried by the foregoing IP bearer system, and the foregoing network resource allocation table And a network resource allocation table of the IP bearer system, which records resources allocated to the specific service on the link inside the switching function unit and the link connecting the switching function unit.

Further, the switching function unit is divided into two types: a switching function unit and an intermediate switching function unit connected to the service network, wherein the intermediate switching function unit is connected to the switching function unit connected to the service network; the switching function unit is internally Non-blocking switching, non-blocking forwarding between logical interfaces; and the above-mentioned switching function unit is connected to a resource-isolated logical network through a logical link, where the logical network is the specific network resource, and the logical network carries a specific service, and It can be planned and managed independently of the logical network of other services.

Further, the logical interface is a logical interface having a resource attribute, which is obtained by dividing the physical interface by the switching function unit according to the resource scheduling and allocation rule delivered by the resource control function unit; and the logical network is controlled by the resource. Functional unit control and management.

Further, the resources in the foregoing resource scheduling and isolation include internal bandwidth resources and interface bandwidth resources.

Further, the service flow requirement table may be manually configured by the network administrator or automatically imported by the service network; the network resource allocation table may be manually calculated by the network administrator or calculated by the resource control function unit according to the service flow requirement table.

To achieve the above, the method for implementing the QoS-guaranteed multi-service IP bearer system of the present invention includes the following steps: Step 1: The network administrator manually configures the service flow requirement table or automatically imports the service flow requirement table by the service network;

Step 2: The network administrator manually calculates the network resource allocation table or the resource control function unit. Generating a network resource allocation table according to the foregoing service flow requirement table;

Step 3: The resource control function unit sends a resource scheduling and allocation rule to the switching function unit according to the foregoing network resource allocation table, and delivers a mapping relationship between the resource and the service to the switching function unit.

Step 4: The switching function unit divides the physical interface into a plurality of logical interfaces having resource attributes according to the resource scheduling and allocation rules delivered by the resource control function unit, and forwards the service traffic with the specific identifier on the logical interface;

Step 5: The switching function unit measures the utilization of the logical link bandwidth, and reports it to the resource control function unit actively or passively;

Step 6: The resource control function unit collects and monitors resource utilization of the logical network, and dynamically allocates network resources between the logical networks, where the network resource is a logical link bandwidth.

Further, in the foregoing step 2, the network administrator or the resource control function unit calculates the foregoing network resource allocation table to include at least the following steps:

Step 201: The network administrator or the resource control function unit calculates, according to the foregoing service flow requirement table, a bandwidth required for the service traffic of the corresponding service on one physical link;

Step 202: The network administrator or the resource control function unit calculates the network resource to be allocated according to the light load factor of the foregoing logical network, that is, the network resource that needs to be allocated - the bandwidth required for the traffic of the corresponding service on one physical link.轻 Light load factor of the logical network, where the network resource is the logical link bandwidth.

Further, the network resource allocation table is manually calculated by the network administrator, or is generated by the resource control function unit according to the traffic demand table, and the specific logical network is generated. The resource allocation meets the requirements of a specific service, and the resource allocation is only related to the demand of the service; the light load factor of the logical network is determined by the QoS required by the specific service carried by the logical network; that is, if the required QoS is higher, The lighter load factor is smaller; if the required QoS is lower, the light load factor is larger.

Further, in the above step 4, each of the foregoing logical interfaces is associated with a forwarding queue, wherein the forwarding queue is given a certain polling priority and bandwidth, and the logical interface is associated with a specific class of service (Class of Service) ( Hereinafter referred to as the CoS, the differentiated service code point (DSCP) or the multi-protocol label switch (hereinafter referred to as MPLS), and the logical interface only forwards CoS and DSCP. Or the service traffic of the MPLS EXP identifier; the service traffic with the specific identifier is the service traffic of the above-mentioned EXP identifier with CoS, DSCP or MPLS.

Further, the above step 4 further includes the following steps:

Step 401: The switching function unit connected to the service network classifies the service traffic according to the mapping relationship between the resource and the service delivered by the resource control function unit, according to the CoS, DSCP or MPLS EXP identifier, according to the IP/MPLS technology. The routing is routed to the physical interface, and is forwarded on the logical interface corresponding to the CoS, DSCP, or MPLS EXP;

Step 402: The intermediate switching function unit receives the service traffic forwarded from the foregoing switching function unit connected to the service network, and routes the route to the physical interface according to the foregoing IP/MPLS technology, and the logic corresponding to the EXP of the CoS, DSCP or MPLS. Forwarded on the interface.

Further, in the foregoing step 6, the method for dynamically allocating network resources between the logical network by the resource control function unit includes: Step 601: The resource control function unit calculates the service traffic on the logical link according to the logical link bandwidth utilization status reported by the switching function unit, where the service traffic on the logical link is the bandwidth currently required by the corresponding service.

Step 602: The resource control function unit calculates the network resource to be allocated according to the light load factor of the foregoing logical network, that is, the network resource to be allocated = the traffic flow on the logical link, the light load factor of the logical network, where the network The resource is the logical link bandwidth.

Further, in the foregoing step 6, the method for dynamically allocating network resources between the logical network by the resource control function unit further includes:

Step 611: The resource control function unit calculates the service traffic on the current logical link according to the logical link bandwidth utilization status reported by the switching function unit, and predicts the logical link in the future period by using a prediction algorithm according to the prediction model extracted by the historical data. The maximum service traffic on the logical link, where the maximum service traffic on the logical link in the future is the bandwidth required by the corresponding service in the future;

Step 612: The resource control function unit calculates the required allocated network resource according to the light load factor of the foregoing logical network, gp: network resource to be allocated = maximum service traffic on the logical link in a future period ÷ light load of the logical network Factor, where the network resource is the logical link bandwidth.

Further, in the above step 611, the foregoing prediction algorithm includes at least an Auto Regressive Moving Average (ARMA) algorithm, a neural network algorithm or a genetic algorithm.

Further, when the resource control function unit dynamically allocates network resources between logical networks, if the service traffic of all services increases, the bandwidth required for the corresponding service increases. If the sum of the logical link bandwidths may exceed the physical bandwidth, the resource control function unit adopts a service priority-based control policy to preferentially ensure the bandwidth requirement of the logical network of the service with high priority.

Further, the implementation method of the foregoing multi-service IP bearer system with QoS guarantee further includes the following steps:

Step 7: The resource control function unit accepts the QoS feedback of the service network, and dynamically adjusts the light load factor of the logical network, that is, if the QoS fed back by the service network does not meet the requirement, the resource control function unit will correspond to the light load factor of the logical network. lower.

The utility model provides a multi-service IP bearer system with QoS guarantee and an implementation method thereof, which can isolate mutual interference between multiple services, and can ensure the telecommunication services carried in the rated service traffic range, such as softswitch, 3G, etc. The QoS of the service.

The above and other objects, features, and advantages of the present invention will become more apparent and understood by the appended claims appended claims DRAWINGS

1 is a block diagram of a component of an IP bearer system according to the present invention;

2 is a schematic flow chart of an implementation method according to the present invention;

3 is a schematic diagram of a network according to a preferred embodiment of the present invention;

4 is a schematic diagram of traffic flow monitoring and logical traffic control of a logical interface according to a preferred embodiment of the present invention;

5 is a schematic diagram of constructing a logical network on a physical network in a preferred embodiment of the invention; FIG. 6 is a schematic diagram of a simple dynamic bandwidth reservation according to the present invention; 7 is a schematic diagram of dynamic bandwidth reservation for bandwidth prediction according to the present invention;

8 is a schematic diagram of processing of QoS feedback of a service network according to the present invention;

FIG. 9 is a schematic diagram of end-to-end QoS guarantee according to the present invention. detailed description

The specific implementation of the technical solution of the present invention will be further described in detail below with reference to the accompanying drawings, which are not intended to limit the invention.

1 is a block diagram of a component of an IP bearer system according to the present invention. As shown in FIG. 1, it includes at least a switching function unit, which supports resource scheduling and isolation, and supports matching specific services with specific network resources and utilizing the specific Network resources forward the specific service; and

The foregoing switching function unit includes a router and a switch. The resource control function unit includes a service flow requirement table and a network resource allocation table, where the service flow requirement table is an end-to-end service flow requirement table of the service network to the IP bearer system. Recording, by the foregoing service network, the service traffic and the flow direction carried by the IP bearer system, where the network resource allocation table is a network resource allocation table of the IP bearer system, and recording the internal of the switching function unit and the link connecting the switching function unit Resources allocated for the above specific services.

2 is a schematic flowchart of the implementation method of the present invention. As shown in FIG. 2, the method mainly includes the following steps;

Step 1: The network administrator manually configures the service flow requirement table or automatically imports it from the service network. Business flow demand list;

Step 2: The network administrator manually calculates a network resource allocation table or is configured by the resource control function unit to generate a network resource allocation table according to the foregoing service flow requirement table;

Assume that a service network is connected to an IP bearer network through three gateway devices GW1, GW2, and GW3. The bearer network is composed of switching functional units (R1, R2, R3, and R4) and three Gigabit links. 3 is shown. The service traffic between the gateway devices GW1 and GW2 is converted to two-way 20 Mbps, and the service traffic between the gateway devices GW2 and GW3 is converted to two-way 35 Mbps, and the service traffic between the gateway devices GW3 and GW1 is converted to two-way 15 Mbps.

The service is intended to be mapped to an IP CoS value of 100 in the bearer network.

The service flow requirement table manually configured by the network administrator or automatically imported by the service network is as shown in Table 1. Table 1

The network administrator manual calculation or resource control function unit calculates the bandwidth requirement on the bearer network link according to the above service traffic demand table, as shown in Table 2:

Table 2

Considering the requirements of QoS indicators such as packet loss rate, delay, and jitter of the service, the light load factor of the logical network that selects the service is 0.5. The network administrator manual calculation or resource control function unit automatically generates the generated network resource allocation table as shown in Table 3:

table 3

CoS link bandwidth requirement link bandwidth allocation

Rl -> R4 100 35Mbps 70Mbps

R4 -> R1 100 35Mbps 70Mbps

R2 -> R4 100 55Mbps 110Mbps R4 -> R2 100 55Mbps 110Mbps

R3 -> R4 100 50Mbps 100Mbps

R4 -> R3 100 50Mbps 100Mbps

The resource control function unit sends the resource allocation rule to the switching function units R1, R2, R3 and R4. The traffic interface of the R1->R4 link and the traffic monitoring of the logical interface are shown in Figure 4.

The logical link connects the above switching functional units into a logically isolated logical network, and a schematic diagram of constructing a logical network independently carrying the above services on the physical network is shown in FIG. 5.

The method for the resource control function unit to dynamically allocate network resources between the logical networks is that the resource control function unit calculates the service traffic on the logical link according to the logical link bandwidth utilization status reported by the switching function unit, and then divides the service traffic by the logical network. The light load factor calculates the current logical link bandwidth to be allocated, and the schematic diagram is as shown in FIG. 6; or, the resource control function unit calculates the current logical link according to the logical link bandwidth utilization status reported by the switching function unit. Traffic flow, and based on prediction models extracted from historical data, such as 24-hour traffic flow change rules, etc., using predictive algorithms, such as ARMA algorithms, neural network algorithms, genetic algorithms, etc., to predict the maximum on the logical link in the future period of time. The traffic is then divided by the light load factor of the logical network to calculate the bandwidth of the logical link that needs to be allocated. The schematic diagram is shown in Figure 7.

The bearer network is isolated by logical network and deduced according to empirical values or theory. The service network requires QoS for the bearer network to configure a light load factor for the logical network. However, based on empirical values or theoretical derivation, the light load factor does not necessarily meet the needs of the service network 100%. Therefore, the service network can monitor the QoS of the bearer network and feed back the QoS indicator to the resource control function unit. Adjust the light load factor, and its processing diagram is shown in Figure 8. If the bearer system QoS fed back by the service network does not meet the requirements, the resource control function unit may reduce the light load factor of the corresponding logical network to allocate more resources for the logical network.

The IP bearer system of the present invention provides a QoS guarantee for the service network within the range of the service traffic from the service gateway to the service gateway. In the rated traffic flow range, combined with the access QoS of the service network, the end-to-end QoS guarantee can be achieved. The schematic diagram is shown in Figure 9.

The above is a detailed description of the working principle of the present invention, but this is only a visual example for the purpose of understanding and should not be construed as limiting the scope of the invention. Also, various possible equivalents and substitutions may be made in accordance with the description of the technical solutions of the present invention and the preferred embodiments thereof, and all such changes or substitutions are intended to fall within the scope of the appended claims.

Claims

Rights request

A multi-service internet protocol bearer system with quality of service guarantee, characterized in that it comprises at least:

The switching function unit includes a router and a switch; the resource control function unit includes a service flow requirement table and a network resource allocation table, where

The service flow requirement table is an end-to-end service flow requirement table of the service network to the network protocol bearer system, and records the service traffic and flow direction carried by the network protocol bearer system.

The network resource allocation table is a network resource allocation table of the Internet Protocol bearer system, and records resources allocated to the specific service on the link of the switching function unit and the link connecting the switching function unit.

2. The system of claim 1 wherein

The switching function unit is divided into two types: a switching function unit and an intermediate switching function unit connected to the service network, wherein the intermediate switching function unit is connected to the foregoing switching function unit connected to the service network;

The above switching functional unit is internally non-blocking and has no blocking forwarding between logical interfaces; The switching function unit is connected to a resource-isolated logical network through a logical link, where the logical network is the specific network resource.

3. The system of claim 2 wherein

The logical interface is a logical interface having a resource attribute, which is obtained by dividing the physical interface by the switching function unit according to the resource scheduling and allocation rule delivered by the resource control function unit;

The above logical network is controlled and managed by the above resource control function unit.

4. The system according to claim 1, wherein the resources in the resource scheduling and isolation comprise internal bandwidth resources and interface bandwidth resources.

5. The system of claim 1 wherein

The above service flow requirement table is manually configured by the network administrator or automatically imported by the above service network;

The network resource allocation table is manually calculated by the network administrator or calculated by the resource control function unit according to the service flow requirement table.

A method for implementing a multi-service internet protocol bearer system with service quality assurance, comprising the following steps: Step 1: The network administrator manually configures a service flow requirement table or automatically imports a service flow demand table by the service network;

Step 3: The resource control function unit sends a resource scheduling and allocation rule to the switching function unit according to the foregoing network resource allocation table, and delivers resources and services to the switching function unit. Shooting relationship

The method according to claim 6, wherein in the step 2, the network administrator or the resource control function unit calculates the network resource allocation table to include at least the following steps: Step 201: Network administrator or resource control function unit Calculating the bandwidth required for the service traffic of the corresponding service on one physical link according to the foregoing service flow requirement table;

Step 202: The network administrator or the resource control function unit calculates the network resource to be allocated according to the light load factor of the foregoing logical network, that is, the network resource to be allocated=the bandwidth required for the traffic of the corresponding service on one physical link.轻 Light load factor of the logical network, where the network resource is the logical link bandwidth.

8. The method of claim 7 wherein

The resource allocation of a specific logical network meets the needs of a specific service, and its resource allocation is only related to the demand of the service;

The light load factor of the above logical network is determined by the quality of service required by the specific service carried by the above logical network; that is, if the required quality of service is higher, the light load factor is smaller; The lower the quality of service, the greater the light load factor.

9. The method according to claim 6, wherein in the above step 4, each of the logical interfaces is associated with a forwarding queue, wherein the forwarding queue is given a certain polling priority and bandwidth, and the logical interface Corresponding to a specific service class, a differentiated service code point, or an EX-switched EXP, and only the service traffic with the service class, the service code point, or the multi-protocol label switching EXP identifier is forwarded on the logical interface;

The above-mentioned service traffic with a specific identifier is the above-mentioned service traffic with the service category, the differentiated service code point or the EXP identifier of the multi-protocol label switching.

The method according to claim 9, wherein the step 4 further comprises the following steps: Step 401: The switching function unit connected to the service network controls the mapping relationship between the resource and the service delivered by the resource control function unit according to the resource control function unit. Traffic traffic classification is colored, according to the service category, differentiated service code point or EXP label of multi-protocol label switching, routed to the physical interface according to the Internet Protocol/Multi-Protocol Label Switching technology, in the above service category, differentiated service code point Or forwarding on the logical interface corresponding to the EXP of the multi-protocol label switching;

Step 402: The intermediate switching function unit receives the service traffic forwarded from the foregoing switching function unit connected to the service network, and routes the route to the physical interface according to the foregoing Internet Protocol/Multi-Protocol Label Switching technology, in the service category and the differentiated service code. Point or multi-protocol label switching is performed on the logical interface corresponding to the EXP.

The method according to claim 6, wherein in the step 6 above, the method for dynamically allocating network resources between the logical network by the resource control function unit comprises: Step 601: The resource control function unit is configured according to the switching function unit Evaluated logical link The bandwidth utilization status calculates the service traffic on the logical link, where the service traffic on the logical link is the bandwidth currently required by the corresponding service;

The method according to claim 6, wherein in the foregoing step 6, the method for dynamically allocating network resources between the logical network by the resource control function unit further includes: Step 611: The resource control function unit is configured according to the foregoing switching function. The logical link bandwidth utilization status reported by the unit calculates the service traffic on the current logical link, and uses the prediction algorithm according to the prediction model extracted from the historical data to predict the maximum service traffic on the logical link in the future period, wherein the future time period The maximum service traffic on the internal logical link is the bandwidth required by the corresponding service in the future;

Step 612: The resource control function unit calculates the network resource to be allocated according to the light load factor of the foregoing logical network, and the network resource to be allocated is: the maximum service traffic on the logical link in the future period ÷ the light load factor of the logic network, The network resource is the logical link bandwidth.

The method according to claim 12, characterized in that in the above step 611, the prediction algorithm comprises at least an autoregressive moving average algorithm, a neural network algorithm or a genetic algorithm.

The method according to claim 11 or 12, wherein when the resource control function unit dynamically allocates network resources between logical networks, the resource control function unit adopts a service priority-based control policy, and prioritizes priority The bandwidth requirements of the logical network of the high-level business.

15. A method according to claim 7 or 11 or 12, further comprising the steps of:

Step 7: The resource control function unit receives the service quality feedback of the service network, and dynamically adjusts the light load factor of the logical network, that is, if the service quality fed back by the service network does not meet the requirement, the resource control function unit will correspond to the light of the logical network. The load factor is reduced.