CN100531257C - Control method and system for preventing load from over loading in communication network - Google Patents

Abstract

The invention relates to a control method for preventing load overload in a communication network, comprising: 1) preset the load upper limit of the processor, and determine the control error of the processor load; 3) The system directly rejects the reported call request message according to the call filtering ratio. The present invention also relates to a control system for preventing load overload in a communication network, comprising: a service control call processing unit; an automatic adjustment unit; a PID controller, connected to the automatic adjuster and the service control call processing unit respectively, and controlling the load The PID control operation is performed on the control error, and the operation result is reported to the service control call processing unit. The invention has the advantages of simple realization, unified load evaluation standard and direct control target, so as to solve the problem of system overload and provide services for users to the greatest extent.

Description

Control method and system for preventing load overload in communication network

technical field

The invention relates to the technical field of system overload control, in particular to a control method and system for preventing load overload in a communication network.

Background technique

Like other systems, in the intelligent network system, in order to ensure the long-term stable operation of the system, especially under the condition of large traffic, overload control measures must be taken to achieve maximum performance under system overload conditions. Guaranteed to provide services to users. The following uses an intelligent service as an example for description.

In the intelligent network, the service control point SCP is the core component of the intelligent network. Its main function is to start different service logics according to the call events reported by the service switching point SSP, and send call control instructions to the corresponding service switching point SSP according to the service logic. , so as to realize various intelligent calls. Usually, the service control point SCP is physically a server or a minicomputer.

There are many kinds of intelligent services, such as 200, 201, 800, etc. are all intelligent services, and one SCP can simultaneously load and process multiple such intelligent services. The realization of an intelligent service requires multiple information interactions between the SCP and the SSP, and a complex intelligent service may include dozens or even hundreds of interactions. As shown in Figure 1, it is a recharge process in a mobile intelligent network, including:

Each user uses this intelligent service to communicate with the called party. The service control point SCP needs to continuously process the messages reported by the service switching point SSP, and then send the corresponding information to the service switching point SSP according to the specific service control logic. Control messages and control the whole process of calling. If there are many users using intelligent services at the same time, the service control point SCP will be quite busy or even overloaded. Consequences of system overload can include lost messages, dropped calls that cannot be processed in time, and system crashes. Therefore, in the communication system, measures to prevent equipment overload are essential, and for service control SCP equipment, there are many ways to prevent system overload.

At present, the system overload prevention mechanism is mainly composed of two parts: the evaluation of the system busyness and the means of reducing the system load. The busyness of the evaluation system can select multiple indicators, and the most common indicators have the following: 1) within a period of time, count the number of call attempts reported by the terminal to the service switching point SSP; 2) the length of the message queue accumulation in the system ; 3) The average statistical value of the system's response time to the call attempt message. In this regard, the intelligent network system itself proposes a call gap mechanism (Call Gapping). When using this mechanism, the new call is blocked (rejected) at the service switching point SSP, that is, the new call arrives at the service The control point SCP is previously blocked, so that the potential of the SCP control unit is not wasted, while ensuring that unblocked calls can be carried through to the end.

In the intelligent network system, a lower limit is set for the service control point SCP, that is, the so-called gap time range. Using this gap time range, the service control point reduces the communication service, that is, the service switching point is not allowed to communicate with the service control point if the time interval is less than this time interval. Report a new call. If the service switching point SSP receives a call and sends it to the service control point SCP, then all new calls within this gap time range will be rejected by the SSP and will not be reported to the SCP, and will only be received after the end of this time range The first new call to arrive at the SSP, and the timer is reset, repeating periodically. This mechanism lasts for a certain amount of time D, called the duration. After this duration expires, the SSP can receive all calls again. The length of the duration is determined by the SCP.

It can be seen that the implementation of the call gap mechanism is relatively complicated. The SCP needs to calculate the call gap time and the control duration, and the SSP needs to add a timing mechanism to ensure that only one call is reported within the call gap time and that the control continues to be specified. time. In addition, usually the SSP also processes ordinary calls at the same time, and the ordinary calls do not need to be reported to the SCP. Therefore, the SSP also needs to ensure that the call gap control will not affect the ordinary calls. For ordinary calls, the SSP is like an ordinary exchange, which directly transfers the call to the destination user through the exchange according to the calling number; while the intelligent call requires the SSP to exchange information with the SCP, which requires the SSP to be able to distinguish between ordinary calls and Intelligent service calls, otherwise the SSP may reject ordinary calls according to the call gap mechanism. Therefore, the call gap mechanism requires the cooperation of the SSP, which also increases the complexity of the mechanism implementation.

In addition, the index parameters for evaluating the busyness of the system, the number of call attempts reported by the SSP within a period of time, the length of the message queue accumulation in the system, the average statistical value of the system’s response time to the call attempt messages, etc., the specific values of these indicators How much is considered to be system overload depends on experience and experiments, and different devices have different processing capabilities, so the specific value of the parameters of the overload index should be modified according to the specific device, which is obviously not convenient.

The index parameters for evaluating the system load are basically related to the call services carried by the equipment. They only reflect the working conditions of the system through the phenomena reflected in the call processing process, which are indirect and cannot directly reflect the execution of call control functions. The system resource utilization of the device (such as a server) itself, such as CPU resource usage.

Therefore, the disadvantages of the prior art are: 1. The implementation of the call gap mechanism is relatively complicated and requires the cooperation of multiple devices; 2. The standard for evaluating system overload is more dependent on experience and has a close relationship with the processing capacity of the device itself ;3. The reference data for evaluating the system load is not direct enough, and can only partially reflect the real situation of the system load.

Contents of the invention

The technical problem solved by the present invention is to provide a control method and system for preventing load overload in a communication network. The present invention is simple to implement, unifies the load evaluation standard and directs the control target, so as to solve the problem that the system is under overload conditions and maximizes the benefits for users. Provide services.

In order to solve the above problems, the present invention provides a control method for preventing load overload in a communication network, including:

Presetting the upper limit of the load of the processor, and determining the control error of the processor load according to the difference between the upper limit of the load and the actual occupancy value of the processor;

The control error is calculated by discrete PID control in the control cycle to obtain a call filtering ratio, and the call filtering ratio is the ratio of the number of call requests that the system should reject to the total number of call requests in the cycle;

The system directly rejects the reported call request message according to the call filtering ratio.

The upper limit of the load of the processor is manually set according to the control parameters.

The process of the discrete PID control operation includes: performing proportional, integral and differential operations on the control error of the processor load.

The process of the discrete PID control operation specifically includes:

Its operation formula is:

$\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; Kp</span>}<span\; class="notranslate">\; *</span><span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; E.</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>\frac{\mathrm{<span\; class="notranslate">\; T</span>}}{{\mathrm{<span\; class="notranslate">\; T</span>}}_{\mathrm{<span\; class="notranslate">\; I</span>}}}\underset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}}{\overset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; k</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\mathrm{<span\; class="notranslate">\; E.</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>\frac{{\mathrm{<span\; class="notranslate">\; T</span>}}_{\mathrm{<span\; class="notranslate">\; D.</span>}}}{\mathrm{<span\; class="notranslate">\; T</span>}}<span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; E.</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; E.</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; ]</span>$

Among them, T=Δt is the sampling period, E(k) is the control error at the kth sampling time, E(k-1) is the control error at the k-1th sampling time, and P(k) is the kth time The discrete call filtering ratio output by the controller when sampling, Kp is the proportional coefficient, T _I is the integral time, T _D is the differential time.

The call filtering ratio is the ratio of the number of call requests rejected by the system to the total number of call requests in this period.

The present invention also provides a control system for preventing load overload in a communication network, including:

The service control call processing unit is a functional entity that realizes the service control function, and is used to process various intelligent calls and complete the information interaction with the service switching point;

The automatic adjustment unit is used to receive the value of the actual processor occupancy rate fed back by the service control call processing unit, and compare it with the upper limit of the processor load set by the system to obtain the control error of the processor load;

The PID controller is connected to the automatic adjuster and the service control call processing unit respectively, performs discrete PID control calculation on the control error of the load, and reports the calculation result to the service control call processing unit.

The automatic adjustment unit includes: an arithmetic unit, which is used for subtracting the preset processor CPU load limit from the value of the actual processor occupancy rate.

The discrete PID control operation specifically includes: performing proportional, integral and differential operations on the control error of the processor load.

Compared with the prior art, the present invention has the following beneficial effects: firstly, the present invention is simple to implement, and only one control unit is added in the existing system, and the control unit is designed according to the PID control algorithm for To adjust the system processor occupancy, it only needs to obtain the control error, and then perform few calculations. It does not need to collect and count various call information, and responds quickly to changes in system load. The load control of the service control node SCP does not require the cooperation of the service switching node SSP. Since the SCP returns a call rejection message conforming to the protocol standard for the rejected call, the SSP call processing unit handles the call rejected by the SCP just like a normal call end. The call request does not need to know the active state of the SCP at all; secondly, the control target of the present invention is the most direct, which is the system CPU occupancy rate that directly reflects the system operation; finally, the evaluation standard of the system load is unified and does not depend on empirical data . The evaluation standard of the present invention to the system load is the CPU occupancy rate, and this parameter has a unified value range of 0% to 100% for systems with different performances, and a unified upper limit of the load can be determined for any system without relying on empirical values .

Description of drawings

Fig. 1 is the flowchart of mobile intelligent network charging in the prior art;

Fig. 2 is the control flowchart of the control method for preventing load overload in the communication network of the present invention;

Fig. 3 is the structural block diagram of the control system that prevents load overload in the communication network of the present invention;

Fig. 4 is a structural block diagram of an application example of the present invention.

Detailed ways

The present invention is based on the use of system resources, that is, the occupancy rate of the processor CPU as a reference index and a control target, and designs a simple load proportional integral differential (PID, Proportion Integral Differential) controller, and according to the PID controller The output result directly restricts the call flow of the user. The PID controller first reads the processor CPU load information of the service switching node, and automatically adjusts the load information; then obtains a call filtering ratio through the PID control operation, and reports the call filtering ratio to the service control switching Service control calls in nodes

Finally, the service control call processing unit directly rejects the user call flow reported by the service switching node according to the call filtering ratio value, so as to prevent the system from overloading.

Now the present invention will be further described in conjunction with the accompanying drawings.

Please refer to Fig. 2, which is a flow chart of the control method for preventing load overload in the communication network of the present invention, including:

Step S11: preset the upper limit of the load of the system processor, and determine the control error of the processor load;

Step S12: calculating the control error through PID control within the control period to obtain a call filtering ratio;

Step S13: The system directly rejects the reported call request message according to the passed call filtering ratio.

The upper limit of CPU load of the system processor is manually set according to the control parameter, which refers to the control target of the CPU usage of the system, and its value range may be 1%-100%. The control error of the system processor load is determined according to the difference between the preset processor load upper limit and the actual processor occupancy value (step S11 ). The control error is obtained through the PID control operation in the control cycle to obtain the call filtering ratio. The PID control operation is actually a proportional (Prop-ortion), integral (Integral), and differential (Differential) operation on the control error, The obtained operation result is acted on the controlled system as a control variable, so that the output of the controlled system is consistent with the control target (step S12). Described PID control operation comprises the following steps:

First, the control error is integrated to obtain the output variable of the controller, and the calculation formula is:

$\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; Kp</span>}<span\; class="notranslate">\; *</span><span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; e</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{{\mathrm{<span\; class="notranslate">\; T</span>}}_{\mathrm{<span\; class="notranslate">\; I</span>}}}<span\; class="notranslate">\; *</span><span\; class="notranslate">\; \&Integral;</span>\mathrm{<span\; class="notranslate">\; e</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span>\mathrm{<span\; class="notranslate">\; dt</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; T</span>}}_{\mathrm{<span\; class="notranslate">\; D.</span>}}\frac{\mathrm{<span\; class="notranslate">\; de</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; dt</span>}}<span\; class="notranslate">\; ]</span>$

Among them, P(t) is the controller output signal, e(t) is the controller deviation signal, Kp is the proportional coefficient, T _I is the integral time, T _D is the differential time.

Secondly, since the input and output values in the computer system are not continuous but discrete, the above formula must be discretized, that is, the output variable of the controller is sampled to obtain the discrete call filter ratio, which is discrete The PID formula is:

$\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; Kp</span>}<span\; class="notranslate">\; *</span><span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; E.</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>\frac{\mathrm{<span\; class="notranslate">\; T</span>}}{{\mathrm{<span\; class="notranslate">\; T</span>}}_{\mathrm{<span\; class="notranslate">\; I</span>}}}\underset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}}{\overset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; k</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\mathrm{<span\; class="notranslate">\; E.</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>\frac{{\mathrm{<span\; class="notranslate">\; T</span>}}_{\mathrm{<span\; class="notranslate">\; D.</span>}}}{\mathrm{<span\; class="notranslate">\; T</span>}}<span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; E.</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; E.</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span>$

Among them, T=Δt is the sampling period, E(k) is the input and output error value during the k sampling, E(k-1) is the input and output error value during the k-1 sampling, P( k) is the call filtering ratio output by the controller at the kth sampling.

Since the continuous system CPU usage value cannot be obtained in the computer system, it is necessary to periodically sample the CPU usage rate, and the sampling period is T. At the beginning of each cycle, the system samples the occupancy rate of the processor. In the UNIX operating system, the current occupancy rate of the processor can be obtained through the vmstat command, and then the obtained CPU occupancy rate value is compared with the set CPU load The upper limit is compared to obtain the control error E(k), use this error as the input of the PID controller, perform discrete PID calculation, obtain the call filtering ratio, and report it to the service control call processing unit of the service control node, the call The filtering ratio is the ratio of the number of call requests that the system should reject to the total number of call requests in this period. The service control call processing unit directly rejects the call request message reported by the service switching node SSP according to the reported call filtering ratio in the next control period, so as to prevent the system from overloading.

However, for a call request rejected by the service control node SCP, once the service control call processing unit receives the user's request, it returns a call rejection message immediately, and the service switching node SSP immediately terminates the call of the current user after receiving the rejection message, and the subsequent service control The information interaction between the node SCP and the service switching node SSP and the call control processing of the service control node SCP are canceled, thereby reducing the workload of the service control node SCP and effectively saving the utilization of system resources of the service control node itself potential, thereby reducing the load on the system. That is to say, when the service control node SCP rejects the user's call request, it returns a call rejection message in the protocol standard to the service switching node SSP, and the service switching node SSP knows that the call needs to be terminated after receiving this message.

If the system load has not dropped to the set upper limit after the end of this control period, then the control error E(k) of this period will be accumulated, resulting in an increase in the calculation result of the call filtering ratio P(k), and the next period will be Deny further calls. Repeat this until the system load drops below the set value. At this time, the output of the PID controller will also be maintained within a relatively stable value range, that is, the rejection of the call request by the service control node SCP also reaches a relatively stable ratio. , the system reaches an equilibrium state.

Please also refer to FIG. 3 , which is a structural block diagram of a control system for preventing overload in a communication network according to the present invention; the control system includes: an automatic adjustment unit 32 , a PID controller 33 and a service control call processing unit 34 .

The service control call processing unit 34 is a functional entity that realizes the service control function, and is used to process various intelligent calls and complete information interaction with service switching contacts;

The automatic adjustment unit 32 is used to receive the value of the actual CPU occupancy rate of the processor fed back by the service control call processing unit, and adjust it with the upper limit of the load of the processor CPU to obtain a load control error, and use it as a PID control device input;

The PID controller 33 is connected to the automatic adjustment unit 32 and the service control call processing unit 34 respectively, performs PID calculation on the load control error, and reports the calculation result to the service control call processing unit.

The automatic adjuster 32 includes an arithmetic unit (not shown in FIG. 3 ), which is used to receive the value of the CPU actual occupancy rate fed back by the service control call processing unit, and perform it with the load upper limit of the set system processor CPU. Subtraction operation, so as to get the load control error, and use it as the input variable of the PID controller.

The PID controller 33 is designed according to the PID control operation in the classical control theory and is used to adjust the CPU usage of the system. The PID control operation is as described above, and will not be repeated here. The setting of the input terminal of the PID controller is based on the difference between the preset upper limit of the CPU occupancy rate and the actual CPU occupancy rate, and the difference is the control error of the processor occupancy rate. Proportional, integral and differential operations are performed on the control error of the processor occupancy rate to obtain a discrete call filtering ratio, which is reported to the service control call processing unit 34 .

The service control call processing unit 34 is a subunit in the service control node SCP. The service control call processing unit 34 has a functional entity for realizing specific service control functions. It can handle various intelligent network protocols and complete the communication with the service switching node. Information exchange between SSPs. The information interaction process between the service switching node and the service control node is: the service control call processing unit in the service control node will receive the intelligent call request reported by the service switching point SSP and carried by the standard intelligent network protocol signaling, and According to the user identification and call type contained in the call request, query the relevant information of the user stored in its own database to determine how to control the call flow of the user, and the service control call processing unit will pass the call control command to the user through the protocol The signaling is sent back to the service switching node SSP, and the service switching node SSP decides how to process the user's call request according to the control command of the service control node SCP. During the processing, multiple signaling interactions between the service switching point SSP and the service control node SCP may also occur according to the call requirements. The service control call processing unit will also use a certain method to limit the call requests of the service switching node SSP according to the current system load. In the above process of service information interaction, the service control call processing unit restricts the call request of the service switching node SSP according to the call filtering ratio generated by the PID controller. The call filtering ratio is the current CPU occupancy rate of the system obtained after the system CPU load is subjected to the PID control operation of the PID controller of the present invention, and fed back to the automatic adjustment unit. Usually the SCP server runs on the UNIX operating system, and the UNIX operating system provides function commands for obtaining the CPU usage rate of the system, such as the vmstat command, which outputs a series of current system performance data, and one of the parameters is id, which is the system idle rate, (100 -id) is the current CPU usage of the system. The PID control operation is the same as above, and will not be repeated here.

Please also refer to FIG. 4 , which is a functional block diagram of an application example of the present invention, including: a service switching node 41 , an automatic adjustment unit 32 , a PID controller 33 and a service control node 42 . All interactive signaling between the service switching node SSP and the service control call processing unit in the service control node SCP must pass through the PID controller, and the PID controller only reports the initial request message of each call from the service switching node SSP Control processing is performed, and other call messages that have been allowed to be processed are reported to the service control node SCP.

The control process of the PID controller 33 is: first, the automatic adjustment unit 32 will obtain the actual CPU load information of the system from the service control call processing unit in the service control node 42, and then pass it with the system CPU load upper limit by itself The arithmetic unit adjusts itself, and sends the adjusted control error of the CPU occupancy rate to the PID controller, and the PID controller performs a PID control operation on the received control error, and a call filtering ratio is obtained as a result of the operation, and the It is reported to the service control node 42 . In one control cycle, the PID controller rejects the received initial call request from the service switching node SSP user according to the call filtering ratio obtained by the PID calculation. After a control cycle ends, the automatic adjustment unit 42 reacquires the CPU load information of the SCP, and sends its load information to the PID controller, and the PID controller calculates a new call filtering ratio, and then uses this new ratio value in the following Call filtering is performed within one control cycle.

The above is only a preferred embodiment of the present invention, it should be pointed out that, for those of ordinary skill in the art, without departing from the principle of the present invention, some improvements and modifications can also be made, and these improvements and modifications can also be made. It should be regarded as the protection scope of the present invention.

Claims (7)

1. A control method for preventing load overload in a communication network, characterized in that it comprises: 1) Presetting the upper limit of the load of the processor, and determining the control error of the processor load according to the difference between the upper limit of the load and the actual occupancy value of the processor; 2) The control error is calculated by discrete PID control in the control cycle to obtain a call filtering ratio, and the call filtering ratio is the ratio of the number of call requests that the system should reject to the total number of call requests in the cycle; 3) The system directly rejects the reported call request message according to the call filtering ratio. 2. The control method for preventing system overload in a communication network according to claim 1, wherein the upper limit of the load of the processor is manually set according to control parameters. 3. The control method for preventing system overload in a communication network according to claim 1, wherein the process of the discrete PID control operation includes: performing proportional, integral and differential operations on the control error of the processor load. 4. The control method for preventing system overload in the communication network according to claim 3, wherein the process of the discrete PID control operation specifically includes: Its operation formula is: $\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; Kp</span>}<span\; class="notranslate">\; *</span><span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; E.</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>\frac{\mathrm{<span\; class="notranslate">\; T</span>}}{{\mathrm{<span\; class="notranslate">\; T</span>}}_{\mathrm{<span\; class="notranslate">\; I</span>}}}\underset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}}{\overset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; k</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\mathrm{<span\; class="notranslate">\; E.</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>\frac{{\mathrm{<span\; class="notranslate">\; T</span>}}_{\mathrm{<span\; class="notranslate">\; D.</span>}}}{\mathrm{<span\; class="notranslate">\; T</span>}}<span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; E.</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; E.</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; ]</span>$ Among them, T=Δt is the sampling period, E(k) is the control error at the kth sampling time, E(k-1) is the control error at the k-1th sampling time, and P(k) is the kth time The discrete call filtering ratio output by the controller when sampling, Kp is the proportional coefficient, T _I is the integral time, T _D is the differential time. 5. A control system for preventing overload in a communication network, comprising: The service control call processing unit is a functional entity that realizes the service control function, and is used to process various intelligent calls and complete the information interaction with the service switching point; It is characterized in that it also includes: The automatic adjustment unit is used to receive the value of the actual processor occupancy rate fed back by the service control call processing unit, and compare it with the upper limit of the processor load set by the system to obtain the control error of the processor load; The PID controller is connected to the automatic adjuster and the service control call processing unit respectively, performs discrete PID control calculation on the control error of the load, and reports the calculation result to the service control call processing unit. 6. The control system for preventing load overload in the communication network according to claim 5, wherein the automatic adjustment unit includes: an arithmetic unit, which is used to compare the load limit of the preset processor CPU with the actual occupancy rate of the processor Numerical values are subtracted. 7. The control system for preventing load overload in a communication network according to claim 6, wherein the discrete PID control operation specifically includes: performing proportional, integral and differential operations on the control error of the processor load.

Images