CN105992282B - Method and device for adjusting quality of service parameters - Google Patents

Abstract

the embodiment of the invention provides a method and a device for adjusting service quality parameters, wherein the method comprises the following steps: periodically acquiring the utilization rate of each resource of a network element of a core network; calculating the utilization rate of the core network element according to the utilization rates of the various resources, and judging that the core network element is congested according to the utilization rate of the core network element; calculating service level protocol priorities of different users according to the user levels and the average consumption flow of the different users and the respective weights of the user levels and the average consumption flow, and acquiring the user with a first service level protocol priority, wherein the first service level protocol priority is higher than other service level protocol priorities; quality of service parameters of a user having a first service level agreement priority are adjusted. When network congestion occurs, a core network element adjusts QoS parameters of network connection of a corresponding user; when the network congestion is relieved, the QoS parameters of the users are restored to the original configuration, so that the network resources can be fully utilized.

Description

Method and device for adjusting quality of service parameters

Technical Field

the present invention relates to core network technologies, and in particular, to a method and an apparatus for adjusting quality of service parameters.

Background

In order to ensure the end-to-end isolation between the service channel of the client and the service channel of the conventional China Mobile internet (CMNET), the operator generally accesses the user by using a dedicated Access Point Name (APN) to establish a dedicated bearer tunnel end-to-end in the wireless Access channel, and isolate the dedicated bearer tunnel from other services. Allocating dedicated core network elements and dedicated access point names to users comprises: and the service access of the user is realized by arranging access special gateway equipment and a special firewall at the core network side and butting the access special gateway equipment and the special firewall with a platform of the client through a special transmission special line.

For each user, according to a Service-Level Agreement (SLA), the core network statically configures access parameters of the user for the APN, such as a bandwidth of the APN, the number of accessible users, and the like.

Configuring a special APN for a Universal Subscriber Identity Module (USIM) card used by each front-end device specified by a client, and realizing functions of an independent service flow channel, an independent service charging, a special service management mode and the like through network data configuration of the special APN; meanwhile, through the configuration of the special IP address pool and the deployment of the special firewall security isolation strategy, the isolation of the network of the client and the external network can be realized, the illegal attack of the external network can be reduced to the maximum extent, and the security protection of the telecom level is realized under a Long Term Evolution (LTE) network.

The prior art has the following problems: the access parameters of users are statically configured on a network element of a core network, when the network is congested, the network Quality of a user with a high priority should be preferentially ensured due to different SLAs signed by the user, but the current network access conditions with different Service levels cannot be adapted, that is, the current static configuration mode of Quality of Service (QoS) can only permanently fix the QoS of the user, and cannot dynamically adjust and recover the QoS of the user with different priority levels in real time according to the network congestion condition.

disclosure of Invention

The technical problem to be solved by the present invention is to provide a method and a device for adjusting QoS parameters, which solve the problem that in the prior art, a QoS static configuration mode can only permanently fix QoS of users, but cannot dynamically adjust and restore QoS of users with different priorities in real time according to network congestion conditions.

In order to solve the above technical problem, an embodiment of the present invention provides a method for adjusting a quality of service parameter, which is applied to a core network element, and the method includes: periodically acquiring the utilization rate of each resource of a network element of a core network; calculating the utilization rate of the core network element according to the utilization rates of the resources, and judging that the core network element is congested according to the utilization rate of the core network element; calculating service level protocol priorities of different users according to user levels and average consumption flows of the different users and respective weights of the user levels and the average consumption flows, and acquiring a user with a first service level protocol priority, wherein the first service level protocol priority is higher than other service level protocol priorities; quality of service parameters of a user having a first service level agreement priority are adjusted.

in the method, before adjusting the quality of service parameters of the user having the first service level agreement priority, the method comprises: and the core network element receives a first signaling sent by a policy and charging control system (PCC), wherein the first signaling contains information for adjusting the service quality parameter.

in the method, periodically acquiring the resource utilization rates of the network elements of the core network includes: and periodically acquiring parameters related to the CPU utilization rate, the memory utilization rate and the network link bandwidth utilization rate of the core network element.

In the method, calculating the utilization rate of the core network element according to the resource utilization rates, and judging that the core network element is congested according to the utilization rate of the core network element includes: calculating the CPU utilization rate of the node control board card, the CPU utilization rate of the session control board card and the CPU utilization rate of the user panel board card, calculating the memory utilization rate of a network element of a core network, and calculating the bandwidth utilization rate of a network link; carrying out weighting operation twice on each utilization rate to obtain the utilization rate of the core network element; and comparing the utilization rate with a utilization rate threshold, and if the utilization rate is greater than the utilization rate threshold, judging that the network element of the core network is congested.

in the method, calculating the CPU utilization rate of the node control board card, the CPU utilization rate of the session control board card and the CPU utilization rate of the user board card comprises the following steps: the CPU utilization rate of the node control board card is the average value of the CPU utilization rates of the processing modules of all the node control board cards, and the node control board card is responsible for the overall control of the nodes; the CPU utilization rate of the session control board card is equal to the average value of the CPU utilization rates of the processing modules of all the session control board cards, and the session control board card is responsible for controlling the load of the user; the CPU utilization of the user plane board is equal to the average of the CPU utilization of the processing modules of the user plane boards, and the user plane board is responsible for transmitting user data.

in the method, calculating the memory utilization rate of the network element of the core network comprises: the memory utilization rate of the node control board card is equal to the average value of the memory utilization rates of the processing modules of all the node control board cards; the memory utilization rate of the session control board card is equal to the average value of the memory utilization rates of the processing modules of all the session control board cards; the memory utilization rate of the user plane card is an average value of the memory utilization rates of the processing modules of the user plane cards.

in the method, calculating the bandwidth utilization of the network link comprises: the network link bandwidth utilization ratio is the bandwidth occupied by actual traffic/the total link bandwidth of the device, wherein the total link bandwidth of the device is obtained according to whether the user panel card ports are full duplex, the number of physical connection lines, whether the physical connection lines are full duplex, and whether link binding is configured.

In the method, performing two weighted operations on each utilization rate to obtain the utilization rate of the core network element comprises: the weights of the board cards in the first weighting operation are respectively obtained as follows: the node control board card is 0.5, the session control board card is 0.35, and the user plane board card is 0.15; calculating the CPU utilization rate and the memory utilization rate of the network element by using the CPU utilization rate and the memory utilization rate of the board card through first weighting; the weights of the CPU utilization rate and the memory utilization rate of the network element obtained in the second weighting are respectively: the CPU utilization rate is 0.55, and the memory utilization rate is 0.45; and calculating the utilization rate of the core network element through the second weighted operation by utilizing the CPU utilization rate and the memory utilization rate of the network element.

in the method, calculating the service level agreement priorities of different users through weighting operation according to the user levels and the average consumption flow of the different users and the respective weights of the user levels and the average consumption flow comprises the following steps: the service LEVEL agreement priority SLA SCORE ═ LEVEL ═ WEIGHTLEvel + TRAFFIC @ WEIGHTtraffic of the user; LEVEL is user grade, and is obtained from user subscription data of a home subscriber server; WEIGHTLEvel is the weight that a user of the corresponding user level has; TRAFFIC is the average consumption flow of a user in m months, and is obtained through a service operation support system: TRAFFIC ═ AVERAGE (N-1 month flow number, N-2 month flow number, N-3 month flow number, … …, N-m month flow number), N is the current month; WEIGHTtrafic is the weight that different average consumed traffic has.

In the method, adjusting the quality of service parameter of the user having the first service level agreement priority comprises: adjusting the service quality parameters including QCI, ARP, MBR, APN-AMBR and UE-AMBR; QCI, quality of service class indicator, is an indicator configured by the network side; ARP, distributing and reserving priority, deciding whether to accept or reject the request for establishing the load bearing or adjusting the request under the condition of resource limitation, and not influencing the scheduling priority; MBR, maximum bit rate, wherein the MBR comprises APN-AMBR and UE-AMBR; APN-AMBR, wherein the aggregated maximum bit rate of all non-guaranteed code rate bearers in all public data network connections associated with an access point name is a subscription parameter in a home subscriber server; UE-AMBR, the aggregate maximum bit rate of all non-guaranteed code rate bearers of the client, and the UE-AMBR processing of the uplink and downlink is executed on the eNodeB.

an apparatus for adjusting quality of service parameters, comprising: the data acquisition unit is used for periodically acquiring the utilization rate of each resource of the network element of the core network; the congestion judging unit is used for calculating the utilization rate of the core network element according to the utilization rates of the various resources and judging that the core network element is congested according to the utilization rate of the core network element; the priority unit is used for calculating the service level protocol priorities of different users according to the user levels and the average consumption flow of the different users and the respective weights of the user levels and the average consumption flow, and acquiring the user with a first service level protocol priority, wherein the first service level protocol priority is higher than other service level protocol priorities; and the parameter adjusting unit is used for adjusting the service quality parameters of the users with the first service level agreement priority.

In the device, the data acquisition unit comprises: the CPU utilization rate module is used for acquiring the CPU utilization rate of the network element of the core network; the memory utilization rate module is used for acquiring the memory utilization rate; and the link module is used for acquiring parameters required by the bandwidth utilization rate of the network link.

in the apparatus, the congestion determining unit includes: the utilization rate calculation module is used for calculating the CPU utilization rate of the node control board card, the CPU utilization rate of the session control board card, the CPU utilization rate of the user panel board card, the memory utilization rate of a network element of a core network and the bandwidth utilization rate of a network link; the utilization rate calculation module is used for carrying out weighting operation twice on each utilization rate to obtain the utilization rate of the core network element; and the utilization rate threshold comparison module is used for comparing the utilization rate with a utilization rate threshold, and if the utilization rate is greater than the utilization rate threshold, judging that the network element of the core network is congested.

In the apparatus, the priority unit includes: the SLA priority calculating module is used for calculating the service LEVEL agreement priority SLA SCORE + TRAFFIC WEIGHTtraffic of the user; LEVEL is user grade, and is obtained from user subscription data of a home subscriber server; WEIGHTLEvel is the weight that a user of the corresponding user level has; TRAFFIC is the average consumption flow of a user in m months, and is obtained through a business operation support system BOSS: TRAFFIC ═ AVERAGE (N-1 month flow number, N-2 month flow number, N-3 month flow number, … …, N-m month flow number), N is the current month; WEIGHTtrafic is the weight that different average consumed traffic has.

in the apparatus, the parameter adjusting unit includes: the client perception index improving module is used for adjusting service quality parameters including QCI, ARP, MBR, APN-AMBR and UE-AMBR; QCI, quality of service class indicator, is an indicator configured by the network side; ARP, distributing and reserving priority, deciding whether to accept or reject the request for establishing the load bearing or adjusting the request under the condition of resource limitation, and not influencing the scheduling priority; MBR, maximum bit rate, wherein the MBR comprises APN-AMBR and UE-AMBR; APN-AMBR, wherein the aggregated maximum bit rate of all non-guaranteed code rate bearers in all public data network connections associated with an access point name is a subscription parameter in a home subscriber server; UE-AMBR, the aggregate maximum bit rate of all non-guaranteed code rate bearers of the client, and the UE-AMBR processing of the uplink and downlink is executed on the eNodeB.

the technical scheme of the invention has the following beneficial effects: periodically collecting each performance index to calculate the congestion condition of a core network element, and adjusting the QoS parameter of the network connection of a corresponding user when the network congestion occurs in the core network element; when the network congestion is relieved, the QoS parameters of the users are restored to the original configuration, so that the network resources can be fully utilized.

Drawings

FIG. 1 is a diagram of a system architecture for statically configuring user parameters;

Fig. 2 is a schematic diagram showing the architecture of the system and the relationship with the mobile communication network;

FIG. 3 is a flow chart illustrating a method for dynamically adjusting quality of service parameters;

FIG. 4 shows a schematic view of the working principle;

FIG. 5 is a schematic diagram illustrating a network resource utilization determination;

Fig. 6 shows a diagram of the download rate of a high priority user over that of a normal user.

Detailed Description

in order to make the technical problems, technical solutions and advantages of the present invention more apparent, the following detailed description is given with reference to the accompanying drawings and specific embodiments.

As shown in fig. 1, a Policy and Charging Control (PCC) system deployment is interfaced with an LTE Electronic Product Code (EPC) network, and maximum bandwidth and traffic carrying capacity for a given user or service is formed by using a combination of Maximum Bit Rate (MBR), Allocation and Retention Priority (ARP) and quality of service Class Identifier (QCI). And dynamically adjusting and recovering the QoS of the users with different priorities in real time according to the network congestion condition.

as shown in fig. 2, in a system architecture for implementing the method and apparatus provided by the embodiment of the present invention, the system in which the method and apparatus provided by the embodiment of the present invention are located is other networking parts of the mobile communication network. The System is butted with a Business Operation Support System (BOSS) through a Simple Object Access Protocol (SOAP) interface, butted with a Policy and Charging Enforcement Function (PCEF) through a File Transfer Protocol (FTP) interface, butted with a Home Subscriber Server (HSS) through a Lightweight Directory Access Protocol (LDAP) interface and used for acquiring required network element index data; and issuing a QoS (quality of service) adjusting instruction to the mobile communication network by interfacing with a Policy and Charging Rules Function (PCRF).

an embodiment of the present invention provides a method for adjusting a quality of service parameter, as shown in fig. 3, which is applied to a core network element, and includes:

Step 301, periodically obtaining each resource utilization rate of a core network element (GGSN/PGW);

step 302, calculating the utilization rate of the core network element according to the utilization rates of the resources, and judging that the core network element is congested according to the utilization rate of the core network element;

Step 303, calculating Service level agreement priorities (SLA priorities, Service-level agreement priorities) of different users according to user levels and average consumption flows of the different users and respective weights of the user levels and the average consumption flows, and acquiring a user with a first Service level agreement priority, wherein the first Service level agreement priority is higher than other Service level agreement priorities;

step 304, adjusting the quality of service parameters of the user having the first service level agreement priority.

by applying the technology provided by the embodiment of the invention, the congestion condition of the network element of the core network is calculated by periodically collecting the data of each performance index, and the network element of the core network adjusts the service quality parameters of the network connection of the user with the high service level protocol priority when the network congestion comes, so as to improve the customer perception index of the user with the first service level protocol priority; when the network congestion is relieved, the service quality parameters of the users are restored to the original configuration, so that the network resources can be fully utilized.

as shown in fig. 4, the method for dynamically adjusting qos parameters is applied to a core network element, and therefore, in the process of solving the technical problem of the present application, the working principle of the core network element includes:

step 401, periodically obtaining the utilization rate of each resource of a core network element;

Step 402, judging the network resource utilization condition through indexes such as core network element CPU, memory utilization rate, network link bandwidth utilization rate conformity and the like, and judging whether the network resource utilization condition meets the trigger condition, if not, turning to step 403, and if so, turning to step 404; and if the triggering condition is not met, judging that the core network element is not congested according to the utilization rate of the core network element.

step 403, if the trigger condition is not met, recovering the dedicated APN QoS parameter or not executing any action, and continuing to periodically acquire the resource utilization rates of the core network elements;

step 404, when the trigger condition is satisfied, the update configuration is issued according to the QoS standard, and the update configuration is used for updating the dedicated APN QoS parameter of the node.

each resource utilization rate of the core network element comprises state information such as resource utilization rate and the like, and specifically comprises core network element CPU utilization rate, memory utilization rate and network link bandwidth utilization rate; the QoS parameters comprise MBR, ARP, QCI and other parameters; the client perception indicators include download rate, response delay, etc.

therefore, the periodically acquiring resource utilization rates of the core network elements includes periodically acquiring CPU utilization rate, memory utilization rate and network link bandwidth utilization rate of the core network elements. Then, judging whether the network element of the core network is congested or not by adopting an algorithm based on the utilization rate of each resource comprises the following steps: and judging the network resource utilization condition by adopting various resource utilization rates such as the coincidence of the CPU utilization rate, the memory utilization rate and the network link bandwidth utilization rate of the core network element.

In a preferred embodiment, calculating an overall utilization ratio of the core network element (for short, a utilization ratio of the core network element) according to the resource utilization ratios, and determining that the core network element is congested according to the utilization ratio of the core network element includes: calculating the CPU utilization rate of the node control board card, the CPU utilization rate of the session control board card and the CPU utilization rate of the user panel board card, calculating the memory utilization rate of a network element of a core network, and calculating the bandwidth utilization rate of a network link;

Calculating the utilization rate of the network element of the core network through two times of weighting operation;

and comparing the utilization rate with a utilization rate threshold, and if the utilization rate is greater than the utilization rate threshold, judging that the network element of the core network is congested.

as shown in fig. 5, the board card of the GGSN/PGW device includes: control panel integrated circuit board (C-PIC, Control Plane) and User's panel integrated circuit board (U-PIC, User Plane), Control panel integrated circuit board divide into Node Control (NC-PIC, Node Control) integrated circuit board and conversation (SC-PIC, Session Control) integrated circuit board, and the responsibility of all kinds of board cards in equipment operation is different:

The Node Control board card is responsible for the overall Control of the Node;

The Session Control board card is responsible for controlling the user load;

The U-PIC card is responsible for transmitting data of a user.

the CPU utilization of the three types of boards is not a complete positive correlation, and when the CPU utilization of the control board is high, it is possible that the CPU utilization of the user board is low, and vice versa. Therefore, the CPU utilization of the three types of boards are considered separately, rather than taking the average as the CPU utilization of the device as a whole.

Currently, each board card is a multi-core CPU, the CPU utilization rate of a node control board card is an average value of the CPU utilization rates of the processing modules of each node control board card, and the node control board card is responsible for the overall control of a node;

the CPU utilization rate of the session control board card is equal to the average value of the CPU utilization rates of the processing modules of all the session control board cards, and the session control board card is responsible for controlling the load of the user;

the CPU utilization of the user plane board is equal to the average of the CPU utilization of the processing modules of the user plane boards, and the user plane board is responsible for transmitting user data.

In a preferred embodiment, in the process of calculating the CPU utilization of the network element of the core network, the method for calculating the CPU utilization of one control panel card or one user panel card includes:

The CPU utilization of any board is MAX (the utilization of CORE1, the utilization of CORE2, the utilization of CORE3, the utilization of CORE4, the utilization of … …, and the utilization of COREn). Therefore, the temperature of the molten metal is controlled,

Calculating the CPU utilization rates of the node control board card NC-PIC, the session board card SC-PIC and the user panel board card U-PIC:

NC-PIC CPU utilization ═ AVERAGE (CPU utilization of NC-PIC1, CPU utilization of NC-PIC2, … …, CPU utilization of NC-PICn);

Calculating the CPU utilization rate of the Session Control C-PIC card:

SC-PIC CPU utilization ═ AVERAGE (CPU utilization of SC-PIC1, CPU utilization of SC-PIC2, … …, CPU utilization of SC-PICn);

calculating the CPU utilization rate of the User Plane PIC card:

U-PIC CPU utilization ═ AVERAGE (CPU utilization of U-PIC1, CPU utilization of U-PIC2, … …, CPU utilization of U-PICn).

in a preferred embodiment, the memory utilization of the network elements of the core network is calculated. As shown in fig. 5, the board card of the GGSN/PGW device includes: control panel cards (C-PIC, Control Plane) and User panel cards (U-PIC, User Plane), the Control panel cards are divided into Node Control (NC-PIC) cards and Session Control (SC-PIC) cards. The flow between the C-PIC cards is load-shared, and the flow between the U-PIC cards is load-shared, so that the utilization rate of the memories of all the cards is basically consistent.

The memory utilization rate of the node control board card is equal to the average value of the memory utilization rates of the processing modules of all the node control board cards;

The memory utilization rate of the session control board card is equal to the average value of the memory utilization rates of the processing modules of all the session control board cards;

The memory utilization rate of the user plane card is an average value of the memory utilization rates of the processing modules of the user plane cards.

Calculating the memory utilization rates of the node control board card NC-PIC, the session board card SC-PIC and the user panel board card U-PIC:

NC-PIC memory utilization ═ AVERAGE (NC-PIC1, NC-PIC2, … …, NC-PICn);

Calculating the memory utilization rate of the Session Control C-PIC card:

SC-PIC memory utilization ═ AVERAGE (SC-PIC1, SC-PIC2, … …, SC-PICn);

Calculating the memory utilization rate of the User Plane PIC card:

U-PIC memory utilization ═ AVERAGE (memory utilization of U-PIC1, memory utilization of U-PIC2, … …, memory utilization of U-PICn).

In a preferred embodiment, the network link bandwidth utilization is calculated as shown in FIG. 5. And calculating the total bandwidth of the link of the equipment according to different configuration modes such as whether the U-PIC card ports are in full duplex, the number of the physical connecting lines, whether the physical connecting lines are in full duplex, whether link-group is configured and the like.

calculating the bandwidth utilization rate of a network link: network link bandwidth utilization is the bandwidth occupied by actual traffic/total bandwidth of the device.

in a preferred embodiment, the two weighted weights calculate the utilization of the network elements of the core network, as shown in fig. 5. The weights for obtaining the first weighting of the board cards are respectively as follows: NC-PIC 0.5, SC-PIC 0.35, U-PIC 0.15;

Calculating the CPU utilization rate and the memory utilization rate of the network element by using the CPU utilization rate and the memory utilization rate of the board card through first weighting;

the weights of the CPU utilization rate and the memory utilization rate of the network element obtained in the second weighting are respectively: the CPU utilization rate is 0.55, and the memory utilization rate is 0.45;

and calculating the utilization rate of the core network element by the second weighting by utilizing the CPU utilization rate and the memory utilization rate of the network element.

and comparing the utilization rate with a utilization rate threshold, and if the utilization rate is greater than the utilization rate threshold, judging that the network element of the core network is congested. Without loss of generality, the utilization rate threshold is 90%, and if the utilization rate is always kept below 90%, network element congestion is not caused by network element performance (CPU and memory); if the utilization rate is more than 90%, the network element node is judged to be congested and the resources are insufficient, namely the condition of adjusting the QoS is considered to be met, and the parameters of different users are adjusted according to the algorithm.

According to the user level and the flow use data in the service operation support system, calculating the service level agreement priority (SLA SCORE) of different users, and acquiring the user with the first service level agreement priority in the service level agreement priority. The algorithm for acquiring the SLA SCORE of the user comprises the following steps:

SLA SCORE＝LEVEL*WEIGHTlevel+TRAFFIC*WEIGHTtraffic；

SLA SCORE is the user's service level agreement priority;

LEVEL is the user grade and is obtained from the user subscription data in the HSS;

WEIGHTLEvel is the weight that the user of the corresponding level has;

TRAFFIC is the average consumption flow of a user in m months, and is obtained through a service operation support system: TRAFFIC ═ AVERAGE (N-1 month flow number, N-2 month flow number, N-3 month flow number, … …, N-m month flow number), N is the current month;

WEIGHTtrafic is the weight that different average consumed traffic has;

and sequencing the service level agreement priorities of different users to obtain a first service level agreement priority.

Specifically, the SLA SCORE is a standard for determining a specific level of a user, and for easily explaining the difference of the levels, a gold card user, a silver card user, a bronze card user, and a general user are taken as an example, where the gold card user may be taken as a user having a first service level agreement priority, or, in some application scenarios, the gold card user and the silver card user may be taken as users having the first service level agreement priority together. And (3) dynamically and automatically adjusting the LEVEL of the user every month, and automatically acquiring the LEVEL in any mode without considering the algorithm of the user LEVEL.

WEIGHTLEVEL is the weight that the user at the corresponding level has, as shown in Table 1:

LEVEL	WEIGHTarpu
		Greater than 30	2
10-30	1.5
		5-10	1.2
less than 5	1

TRAFFIC is the average consumption flow of a user in m months, and is acquired by a service operation support system;

TRAFFIC ═ AVERAGE (N-1 month flow number, N-2 month flow number, N-3 month flow number, … …, N-m month flow number), N is the current month;

WEIGHTtrafic is the weight that the corresponding average consumed traffic has, as shown in Table 2:

TRAFFIC	WEIGHTtraffic
		Greater than 15TB	2
2TB-15TB	1.7
		0.5TB-3TB	1.3
Less than 0.5TB	1

According to the user grade and the average consumption flow using condition, SLA SCORE can be calculated, and then the SLA grading of the user is judged as shown in a table 3:

SLA CORE	SLA grading
		greater than 75	gold medal user
50-75	silver card user
		15-50	copper plate user
less than 15	general users

the QoS guarantee criteria for different users in congestion state are as shown in table 4:

SLA grading

QCI

ARP

MBR

gold medal user	6	1	80M
				silver card user	7	2	40M
Copper plate user	8	3	20M
				General users	9	3	10M

in a preferred embodiment, configuration updates are issued to the relevant network elements. And issuing a message to a Policy and Charging Rules Function (PCRF), wherein the PCRF issues a control signaling through a Gx interface, and controls a related network element to adjust QCI, ARP and MBR parameters of the special APN. In other words, the QoS parameter is issued and adjusted through the Gx interface of the PCC system standard, and the core network GGSN/PGW network element receives the issued signaling at the Gx interface.

In a preferred embodiment, adjusting the quality of service parameters of the user having the first service level agreement priority comprises:

Adjusting the service quality parameters including QCI, ARP, MBR, APN-AMBR and UE-AMBR;

QCI, quality of service class indicator, is an indicator configured by the network side;

ARP, distributing and reserving priority, deciding whether to accept or reject the request for establishing the load bearing or adjusting the request under the condition of resource limitation, and not influencing the scheduling priority;

MBR, maximum bit rate, wherein the MBR comprises APN-AMBR and UE-AMBR;

APN-AMBR, wherein the aggregated maximum bit rate of all non-guaranteed code rate bearers in all public data network connections associated with an access point name is a subscription parameter in a home subscriber server;

UE-AMBR, the aggregate maximum bit rate of all non-guaranteed code rate bearers of the client, and the UE-AMBR processing of the uplink and downlink is executed on the eNodeB.

Specifically, the QCI is mainly configured by the network side and includes parameters: resource Type, scheduling Priority (Priority), pdb (Packet Delay bucket), Packet Error Rate, and the like;

ARP: the main purpose is to decide whether to accept or reject the bearer establishment or adjustment request under the condition of resource limitation, mainly for use by the admission module, without affecting the scheduling priority.

MBR: a maximum bit rate including APN-AMBR and UE-AMBR;

APN-AMBR: -per APN aggregation Maximum Bit Rate, APN-AMBR being a subscription parameter in the HSS, -aggregated Maximum Bit Rate of all non-GBR bearers in all Public Data Network (PDN) connections associated with one APN; the uplink APN-AMBR processing is executed on the UE and the PDN GW, and the downlink APN-AMBR processing is executed on the PDN GW. non-GBR refers to the requirement that traffic (or bearers) be subjected to a reduced rate in the event of network congestion.

UE-AMBR: per UE Aggregate Maximum Bit Rate, Aggregate Maximum Bit Rate for all non-GBR bearers in the UE, uplink and downlink UE-AMBR processing is performed at the eNodeB.

in a preferred embodiment, the resources are scarce, and the manner of mitigating includes: QoS parameter recovery: the indicators of CPU, memory and bandwidth are monitored periodically (5 minutes) and if the QoS adjusted parameters are found to no longer be met, the QoS parameters are restored to the standard state.

in an application scenario, the indexes of a core network element within 30 minutes are as shown in table 5:

TABLE 5

Index (I)	board card/network element	5 minutes	10 minutes	15 minutes	20 minutes
						CPU(％)	NC-PIC1	65	65	65	65
	SC-PIC1	78	78	78	78
							SC-PIC2	76	76	76	76
	U-PIC1	55	88	88	86
							U-PIC2	56	88	88	85
	U-PIC-3	55	87	87	85
							U-PIC4	56	89	89	87
Memory(％)	NC-PIC1	55	55	55	55
							SC-PIC1	54	54	54	54
	SC-PIC2	56	56	56	56
							U-PIC1	76	87	86	88
	U-PIC2	77	88	85	88
							U-PIC-3	77	87	85	88
	U-PIC4	77	87	87	86
						Bandwidth(％)	Node	65	87	86	88

step a, according to the network element congestion algorithm, the CPU utilization, memory utilization, and utilization of the network element in 5 minutes, 10 minutes, 15 minutes, and 20 minutes can be calculated as shown in table 6:

TABLE 6

	5 minutes	10 minutes	15 minutes	20 minutes
					Network element CPU utilization (%)	68	73	73	72
Network element memory utilization (%)	58	60	60	60
					Utilization (%)	63	67	67	67

The utilization remains below the utilization threshold of 90% at all times, so network element performance (CPU and memory) does not cause network element congestion.

And step B, according to the network element congestion algorithm, looking up the index of the network element recorded in the table 5 in the past 30 minutes, and thus, the bandwidth utilization rate of continuous three-time sampling in 10 minutes, 15 minutes and 20 minutes is more than 85 percent, so that the network element congestion is caused by the bandwidth performance, and the QoS parameter of the user needs to be adjusted at the 20 th minute.

step C, user rating and traffic usage for three months (m ═ 3) for one user as in table 7:

	one month	february of February	March
				LEVEL	40	35	38
TRAFFIC	12T	10T	10T

According to the SLA scoring algorithm, the SLA SCORE of the user can be calculated:

SLA SCORE＝(40+35+38)/3*2+(12+10+10)/3*1.7＝93；

According to the SLA CORE and the corresponding table of the SLA grades, the user can be judged to be a gold card user.

Step D, because network element congestion has occurred, the QoS of the gold user is guaranteed, and the original QCI is 9, ARP is 3, MBR is 10M, and the QoS is promoted to QCI is 6, ARP is 1, and MBR is 80M.

As shown in fig. 6, after QoS is guaranteed according to the statistical data, the download rate of the high priority user (the first service level agreement priority user, i.e. the gold user) is better than that of the normal user. Where the download rate is a customer perception indicator for the user.

the embodiment of the invention also provides a device for adjusting the service quality parameters, which comprises:

the data acquisition unit is used for periodically acquiring the utilization rate of each resource of the network element of the core network;

the congestion judging unit is used for calculating the utilization rate of the core network element according to the utilization rates of the various resources and judging that the core network element is congested according to the utilization rate of the core network element;

The priority unit is used for calculating the service level protocol priorities of different users according to the user levels and the average consumption flow of the different users and the respective weights of the user levels and the average consumption flow, and acquiring the user with a first service level protocol priority, wherein the first service level protocol priority is higher than other service level protocol priorities;

And the parameter adjusting unit is used for adjusting the service quality parameters of the users with the first service level agreement priority.

in a preferred embodiment, the data acquisition unit comprises:

the CPU utilization rate module is used for acquiring the CPU utilization rate of the network element of the core network;

the memory utilization rate module is used for acquiring the memory utilization rate;

and the link module is used for acquiring parameters required by the bandwidth utilization rate of the network link.

In a preferred embodiment, the congestion judging unit includes:

The utilization rate calculation module is used for calculating the CPU utilization rate of the node control board card, the CPU occupation of the session control board card, the CPU utilization rate of the user panel board card, the memory utilization rate of a network element of a core network and the bandwidth utilization rate of a network link;

The utilization rate calculation module is used for calculating the utilization rate of the core network element by the weighted weights of the two times;

And the utilization rate threshold comparison module is used for comparing the utilization rate with a utilization rate threshold, and if the utilization rate is greater than the utilization rate threshold, judging that the network element of the core network is congested.

In a preferred embodiment, the priority unit comprises:

SLA priority calculation module to calculate

SLA SCORE＝LEVEL*WEIGHTlevel+TRAFFIC*WEIGHTtraffic；

SLA SCORE is a criterion for determining the user's specific LEVEL, LEVEL is the user LEVEL, obtained from the user subscription data in HSS, weight is the weight that the user of the corresponding LEVEL has,

TRAFFIC is the average consumption flow of a user in m months, and is acquired by a service operation support system;

TRAFFIC ═ AVERAGE (N-1 month TRAFFIC, N-2 month TRAFFIC, N-3 month TRAFFIC, … …, N-m month TRAFFIC), N is the current month, and WEIGHTtrafic is the weight that the corresponding consumed TRAFFIC has;

And sequencing the service level agreement priorities of different users to obtain a first service level agreement priority.

In a preferred embodiment, the parameter adjusting unit includes:

the client perception index improving module is used for adjusting service quality parameters including QCI, ARP, MBR, APN-AMBR and UE-AMBR;

QCI, quality of service class indicator, is an indicator configured by the network side;

ARP, distributing and reserving priority, deciding whether to accept or reject the request for establishing the load bearing or adjusting the request under the condition of resource limitation, and not influencing the scheduling priority;

MBR, maximum bit rate, wherein the MBR comprises APN-AMBR and UE-AMBR;

APN-AMBR, wherein the aggregated maximum bit rate of all non-guaranteed code rate bearers in all public data network connections associated with an access point name is a subscription parameter in a home subscriber server;

UE-AMBR, the aggregate maximum bit rate of all non-guaranteed code rate bearers of the client, and the UE-AMBR processing of the uplink and downlink is executed on the eNodeB.

the advantages after adopting this scheme are: periodically collecting each performance index to calculate the congestion condition, after the core network GGSN/PGW network element receives the first signaling, the network congestion comes, and adjusting the service quality parameter of the network connection of the corresponding user so as to improve the customer perception index of the user with the first service level protocol priority; when the network congestion is relieved, the service quality parameters of the users are restored to the original configuration, so that the network resources can be fully utilized. According to the statistical data, after QoS guarantee is carried out, the downloading rate of the gold medal user is superior to that of the common user.

While the foregoing is directed to the preferred embodiment of the present invention, it will be understood by those skilled in the art that various changes and modifications may be made without departing from the spirit and scope of the invention as defined in the appended claims.

Claims (12)

1. a method for adjusting quality of service parameters is applied to a core network element, and is characterized in that the method comprises the following steps:

Periodically acquiring the utilization rate of each resource of a network element of a core network;

calculating the utilization rate of the core network element according to the resource utilization rates, and judging that the core network element is congested according to the utilization rate of the core network element, wherein the method comprises the following steps: calculating the CPU utilization rate of the node control board card, the CPU utilization rate of the session control board card and the CPU utilization rate of the user panel board card, calculating the memory utilization rate of a network element of a core network, and calculating the bandwidth utilization rate of a network link; carrying out weighting operation twice on each utilization rate to obtain the utilization rate of the core network element, wherein the utilization rate comprises the following steps: the weights of the board cards in the first weighting operation are respectively obtained as follows: the node control board card is 0.5, the session control board card is 0.35, and the user plane board card is 0.15;

calculating the CPU utilization rate and the memory utilization rate of the network element by using the CPU utilization rate and the memory utilization rate of the board card through first weighting;

The weights of the CPU utilization rate and the memory utilization rate of the network element obtained in the second weighting are respectively: the CPU utilization rate is 0.55, and the memory utilization rate is 0.45;

Calculating the utilization rate of the core network element through a second weighting operation by utilizing the CPU utilization rate and the memory utilization rate of the network element;

comparing the utilization rate of the core network element with a utilization rate threshold, and if the utilization rate is greater than the utilization rate threshold, judging that the core network element is congested;

Calculating service level protocol priorities of different users according to user levels and average consumption flows of the different users and respective weights of the user levels and the average consumption flows, and acquiring a user with a first service level protocol priority, wherein the first service level protocol priority is higher than other service level protocol priorities;

Quality of service parameters of a user having a first service level agreement priority are adjusted.

2. The method of claim 1, wherein adjusting the quality of service parameters for the user having the first service level agreement priority comprises:

And the core network element receives a first signaling sent by a policy and charging control system (PCC), wherein the first signaling contains information for adjusting the service quality parameter.

3. The method of claim 1, wherein periodically obtaining resource utilization of network elements of a core network comprises:

And periodically acquiring parameters related to the CPU utilization rate, the memory utilization rate and the network link bandwidth utilization rate of the core network element.

4. The method of claim 1, wherein calculating the CPU utilization of the node control board, the CPU utilization of the session control board, and the CPU utilization of the user board comprises:

the CPU utilization rate of the node control board card is the average value of the CPU utilization rates of the processing modules of all the node control board cards, and the node control board card is responsible for the overall control of the nodes;

the CPU utilization rate of the session control board card is equal to the average value of the CPU utilization rates of the processing modules of all the session control board cards, and the session control board card is responsible for controlling the load of the user;

the CPU utilization of the user plane board is equal to the average of the CPU utilization of the processing modules of the user plane boards, and the user plane board is responsible for transmitting user data.

5. The method of claim 1, wherein calculating a core network element memory utilization comprises:

the memory utilization rate of the node control board card is equal to the average value of the memory utilization rates of the processing modules of all the node control board cards;

The memory utilization rate of the session control board card is equal to the average value of the memory utilization rates of the processing modules of all the session control board cards;

the memory utilization rate of the user plane card is an average value of the memory utilization rates of the processing modules of the user plane cards.

6. the method of claim 1, wherein calculating network link bandwidth utilization comprises:

the network link bandwidth utilization ratio is the bandwidth occupied by actual traffic/the total link bandwidth of the device, wherein the total link bandwidth of the device is obtained according to whether the user panel card ports are full duplex, the number of physical connection lines, whether the physical connection lines are full duplex, and whether link binding is configured.

7. the method of claim 1, wherein calculating the service level agreement priorities of different users through a weighting operation according to the user levels and average consumption traffic of the different users and the respective weights of the user levels and the average consumption traffic comprises:

the service LEVEL agreement priority SLA SCORE ═ LEVEL ═ WEIGHTLEvel + TRAFFIC @ WEIGHTtraffic of the user;

LEVEL is user grade, and is obtained from user subscription data of a home subscriber server;

WEIGHTLEvel is the weight that a user of the corresponding user level has;

TRAFFIC is the average consumption flow of a user in m months, and is obtained through a business operation support system BOSS: TRAFFIC ═ AVERAGE (N-1 month flow number, N-2 month flow number, N-3 month flow number, … …, N-m month flow number), N is the current month;

WEIGHTtrafic is the weight that different average consumed traffic has.

8. the method of claim 1, wherein adjusting quality of service parameters for a user having a first service level agreement priority comprises:

adjusting the service quality parameters including QCI, ARP, MBR, APN-AMBR and UE-AMBR;

QCI, quality of service class indicator, is an indicator configured by the network side;

ARP, distributing and reserving priority, deciding whether to accept or reject the request for establishing the load bearing or adjusting the request under the condition of resource limitation, and not influencing the scheduling priority;

MBR, maximum bit rate, wherein the MBR comprises APN-AMBR and UE-AMBR;

APN-AMBR, wherein the aggregated maximum bit rate of all non-guaranteed code rate bearers in all public data network connections associated with an access point name is a subscription parameter in a home subscriber server;

UE-AMBR, the aggregate maximum bit rate of all non-guaranteed code rate bearers of the client, and the UE-AMBR processing of the uplink and downlink is executed on the eNodeB.

9. an apparatus for adjusting quality of service parameters, comprising:

the data acquisition unit is used for periodically acquiring the utilization rate of each resource of the network element of the core network;

The congestion judging unit is used for calculating the utilization rate of the core network element according to the utilization rates of the various resources and judging that the core network element is congested according to the utilization rate of the core network element;

The priority unit is used for calculating the service level protocol priorities of different users according to the user levels and the average consumption flow of the different users and the respective weights of the user levels and the average consumption flow, and acquiring the user with a first service level protocol priority, wherein the first service level protocol priority is higher than other service level protocol priorities;

The parameter adjusting unit is used for adjusting the service quality parameters of the users with the first service level agreement priority;

The congestion determination unit includes:

The utilization rate calculation module is used for calculating the CPU utilization rate of the node control board card, the CPU utilization rate of the session control board card, the CPU utilization rate of the user panel board card, the memory utilization rate of a network element of a core network and the bandwidth utilization rate of a network link;

The utilization rate calculation module is used for carrying out weighting operation twice on each utilization rate to obtain the utilization rate of the core network element; wherein, the weights of the board cards in the first weighting operation are obtained as follows: the node control board card is 0.5, the session control board card is 0.35, and the user plane board card is 0.15; calculating the CPU utilization rate and the memory utilization rate of the network element by using the CPU utilization rate and the memory utilization rate of the board card through first weighting; the weights of the CPU utilization rate and the memory utilization rate of the network element obtained in the second weighting are respectively: the CPU utilization rate is 0.55, and the memory utilization rate is 0.45; calculating the utilization rate of the core network element through a second weighting operation by utilizing the CPU utilization rate and the memory utilization rate of the network element;

And the utilization rate threshold comparison module is used for comparing the utilization rate of the core network element with a utilization rate threshold, and if the utilization rate is greater than the utilization rate threshold, judging that the core network element is congested.

10. the apparatus of claim 9, wherein the data acquisition unit comprises:

The CPU utilization rate module is used for acquiring the CPU utilization rate of the network element of the core network;

the memory utilization rate module is used for acquiring the memory utilization rate;

And the link module is used for acquiring parameters required by the bandwidth utilization rate of the network link.

11. the apparatus of claim 9, wherein the priority unit comprises:

SLA priority calculation module to calculate

the service LEVEL agreement priority SLA SCORE ═ LEVEL ═ WEIGHTLEvel + TRAFFIC @ WEIGHTtraffic of the user;

LEVEL is user grade, and is obtained from user subscription data of a home subscriber server;

WEIGHTLEvel is the weight that a user of the corresponding user level has;

TRAFFIC is the average consumption flow of a user in m months, and is obtained through a business operation support system BOSS: TRAFFIC ═ AVERAGE (N-1 month flow number, N-2 month flow number, N-3 month flow number, … …, N-m month flow number), N is the current month;

WEIGHTtrafic is the weight that different average consumed traffic has.

12. The apparatus of claim 9, wherein the parameter adjusting unit comprises:

the client perception index improving module is used for adjusting service quality parameters including QCI, ARP, MBR, APN-AMBR and UE-AMBR;

QCI, quality of service class indicator, is an indicator configured by the network side;

ARP, distributing and reserving priority, deciding whether to accept or reject the request for establishing the load bearing or adjusting the request under the condition of resource limitation, and not influencing the scheduling priority;

MBR, maximum bit rate, wherein the MBR comprises APN-AMBR and UE-AMBR;

APN-AMBR, wherein the aggregated maximum bit rate of all non-guaranteed code rate bearers in all public data network connections associated with an access point name is a subscription parameter in a home subscriber server;

UE-AMBR, the aggregate maximum bit rate of all non-guaranteed code rate bearers of the client, and the UE-AMBR processing of the uplink and downlink is executed on the eNodeB.