Method for controlling overload in centralized and distributed operating hybrid switching system

• US 5,513,257 A
• Filed: 11/22/1994
• Issued: 04/30/1996
• Est. Priority Date: 12/15/1993
• Status: Expired due to Term

• Current Assignee: Electronics and Telecommunications Research Institute, Korea Telecommunication Authority
• Original Assignee: ABD Korea Telecommunications Authority, Electronics and Telecommunications Research Institute
• Inventors: Kim, Young S., Yoo, Chan H., Lee, Byung S.
• Primary Examiner(s): Hofsass, Jeffery
• Assistant Examiner(s): Hunter, Daniel S.
• Application Number:
  US08/343,649
• Time in Patent Office: 525 Days
• Field of Search: 379/34, 379/113, 379/134, 379/220, 379/221, 379/279
• US Class Current: 379/221.03
• CPC Class Codes: H04M 3/365 Load metering of control unit
• CPC Class Codes: H04Q 3/54591 Supervision, e.g. fault localisation, traffic measurements, avoiding errors, failure recovery, monitoring, statistical analysis

Abstract

A method for controlling an overload in a hybrid full electronic switching system, capable of maintaining a stable service condition for an overload control interval and realizing a priority control depending on the type of call by sorting an overload occurring in the system into an overload associated with processors adapted to execute services in a centralized manner and an overload associated with processors adapted to execute same services in a distributed manner and thereby automatically calculating numbers of calls to be accepted respectively for control intervals for two kinds of overloads in different control manners. The method includes two different control procedures respectively applied to a case wherein a reduction in overload can be carried out in a processor involving the overload and a case wherein a reduction in overload can be carried out by limiting services to be executed in other processors.

Claims

1. A method for controlling an overload in a full electronic switching system including an operation and maintenance processor adapted to record an accounting and to provide an interface with an operator, a number translating processor adapted to perform a number translation, an internal network processor adapted to connect space switches, and a plurality of access switching processors arranged in a distributed manner around the operation and maintenance processor, the number translating processor and the internal network processor, the method comprising the steps of:

(A) executing a control for an overload caused by an operator by use of an originating call approval value, an input trunk call approval value, and a terminating and output trunk call approval value for calls from the operator to be accepted for a first predetermined time period;

(B) detecting an execution delay condition of each of the processors in the order of priorities respectively assigned to the processors and determining a system overload of each processor on the basis of the result of the detection;

(C) detecting a CPU occupancy rate of each processor at intervals of a second predetermined time period, detecting an execution delay condition of a call processing process at intervals of the first predetermined time period, and determining a traffic overload of the processor on the basis of the detected CPU occupancy rate and the detected execution delay condition; and

(D) executing an overload control depending on a kind of overload generated, as one of an operator overload, a system overload and a traffic overload, in a selected one of the access switching processors associated with the generated overload at intervals of the first predetermined time period.

2. A method in accordance with claim 1, wherein the step (A) comprises the steps of:

(a) receiving, at a traffic and resource management software, the originating call approval value, the input trunk call approval value, and the terminating and output trunk call approval value generated from the operator for the first predetermined time period;

(b) sending the call approval values from the traffic and resource management software to a traffic and resource control software loaded in a selected one of the access switching processors associated with the overload; and

(c) outputting the terminating and output trunk call approval value from the traffic and resource control software.

3. A method in accordance with claim 1, wherein the step (B) comprises the steps of:

(a) receiving, at a traffic and resource management software, a reference execution delay value for each of the processes from the operator in the order of priorities respectively assigned to the processes;

(b) sending the reference execution delay value from the traffic and resource management software to a traffic and resource control software loaded in a selected one of the processors associated with the overload; and

(c) determining a system overload of the selected processor on the basis of the reference execution delay value.

4. A method in accordance with claim 1, wherein the step (C) comprises the steps of:

(a) receiving, at a traffic and resource management software, a CPU reference value from the operator;

(b) sending the CPU reference value from the traffic and resource management software to a traffic and resource control software loaded in a selected one of the processors associated with the overload; and

(c) determining a traffic overload of the selected processor on the basis of the CPU reference value.

5. A method in accordance with claim 3, wherein the step (c) comprises the steps of:

(c-1) if the selected processor is one of the operation and maintenance processor, the number translating processor and the internal network processor, checking at intervals of the second predetermined time period about whether each of the processes of the selected processor respectively having priorities of 0, 1, 2, and 3 has delayed its execution successive times not less than the reference execution delay value, and determining the system overload on the basis of the result of the checking;

(c-2) if the system overload has occurred, informing all of the access switching processors of the occurrence of the system overload to suppress originating calls and input trunk calls in all of the access switching processors; and

(c-3) demanding a decrease in non-traffic approval value of all of the access switching processors when the selected process has delayed for a time not less than the reference execution delay value under the system overload condition;

(c-4) releasing the system overload when no call suppressed indication has been generated in the selected access switching processor while the delay time of the selected process is less than the reference execution delay value under the system overload condition; and

(c-5) demanding an increase in non-traffic approval value of all access switching processors when at least one call suppressed indication has been generated in the selected access switching processor while the delay time of the selected process is less than the reference execution delay value under the system overload condition.

6. A method in accordance with claim 3, wherein the step (c) comprises the steps of:

(c-1) if the selected processor is one of the access switching processors, checking at intervals for the second predetermined time period about whether each of the processes of the selected processor respectively having priorities of 0, 1, 2, and 3 has delayed its execution successive times not less than the reference execution delay value, and determining the system overload on the basis for the result of the checking;

(c-2) if the system overload has occurred, setting a non-traffic approval value at zero and thereby suppressing all originating and input trunk calls;

(c-3) executing a decrease in non-traffic approval value when the selected process has delayed for a time not less than the reference execution delay value under the system overload condition;

(c-4) releasing the system overload when no call suppressed indication has been generated under the system overload condition while the delay time of the selected process is less than the reference execution delay value; and

(c-5) executing an increase in non-traffic approval value when at least one call suppressed indication has been generated while the delay time of the selected process is less than the reference execution delay value under the system overload condition.

7. A method in accordance with claim 4, wherein the step (c) comprises the steps of:

(c-1) if the selected processor is one of the operation and maintenance processor, the number translating processor and the internal network processor, checking at intervals of the second predetermined time period whether the delayed condition of the call processing process detected at intervals of the first predetermined time period has occurred predetermined successive times or has been continued for the second predetermined time period and whether the CPU occupancy rate is not less than the CPU reference value, under a condition that no traffic overload has occurred, and making a determination that a traffic overload has occurred when the delayed condition of the call processing process has occurred predetermined successive times or has been continued for the second predetermined time period while the CPU occupancy rate is not less than the CPU reference value;

(c-2) informing all of the access switching processors of the occurrence of the traffic overload to set a non-traffic overload to set a non-traffic approval value;

(c-3) demanding a decrease in non-traffic approval value of all of the access switching processors when the CPU occupancy rate is not less than the CPU reference value under the traffic overload condition;

(c-4) releasing the traffic overload when no call suppressed indication has been generated in the selected access switching processor predetermined successive times while the CPU occupancy rate is less than the CPU reference value under the traffic overload condition; and

(c-5) demanding an increase in non-traffic approval value of all of the access switching processors when at least one call suppressed indication has been generated in the selected access switching processor.

8. A method in accordance with claim 5 wherein a change of the non-traffic approval value for executing the decrease in non-traffic approval value and the increase in non-traffic approval value is carried out by the steps of:

(a) when the non-traffic approval value is to be changed, setting a minimum approval value at zero, setting a maximum approval value at a sum of the number of successful originating calls and the number of successful input trunk calls respectively generated for 10 seconds, and setting the non-traffic approval value by the minimum approval value;

(b) when the decrease in non-traffic approval value is required, setting a minimum approval value at a value derived by deducting the value resulting from division of the difference between previous minimum and maximum approval values by 2 from the previous minimum approval value, setting a maximum approval value at a value derived by dividing the difference between the newly set minimum approval value and the previous maximum approval value by 2, and setting the non-traffic approval value at the newly set minimum approval value; and

(c) when the increase in non-traffic approval value is required, setting a maximum approval value at a value derived by adding the value resulting from division of the difference between the previous minimum and maximum approval values by 2 to the previous maximum approval value, setting a minimum approval value at a value derived by dividing the difference between the previous minimum approval value and the newly set maximum approval value by 2, and setting the non-traffic approval value at the newly set maximum approval value.

9. A method in accordance with claim 4, wherein the step (c) comprises the steps of:

(c-1) if the selected processor is one of the access switching processors, checking at intervals of the second predetermined time period whether the delayed condition of the call processing process detected at intervals of the first predetermined time period has occurred predetermined successive times or has been continued for the second predetermined time period and whether the CPU occupancy rate is not less than the CPU reference value, under a condition that no traffic overload has occurred, and making a determination that a traffic overload has occurred when the delayed condition of the call processing process has occurred predetermined successive times or has been continued for the second predetermined time period while the CPU occupancy rate is not less than the CPU reference value;

(c-2) initializing a process standby limiting value for the call processing process when the traffic overload has occurred;

(c-3) releasing the traffic overload when no call suppressed indication has been generated predetermined successive times while the CPU occupancy rate is less than the CPU reference value under the traffic overload condition;

(c-4) increasing the process standby limiting value when at least one call suppressed indication has been generated while the CPU occupancy rate is less than the CPU reference; and

(c-5) decreasing the process standby limiting value when the CPU occupancy rate is not less than the CPU reference.

10. A method in accordance with claim 1, wherein the step (D) comprises the steps of:

(d-1) when a call approval value has been determined in accordance with a command from the operator upon the overload control in the selected access switching processor, setting an operator call approval value by the determined call approval value;

(d-2) when an overload has occurred in one of the operation and maintenance processor, the number translating processor and the internal network processor or when a system overload has occurred in the selected access switching processor, calculating a non-traffic originating call approval value by deducting the number of input trunk calls generated for a predetermined interval from a non-traffic approval value, and then setting a non-traffic input trunk call approval value at the non-traffic approval value;

(d-3) when a traffic overload has occurred in the selected access switching processor, calculating a call approval value by adding a process standby limiting value for the call processing process to the number of calls accepted for the predetermined interval, and then deducting the number of standby call processing processes from the value resulting from the addition;

(d-4) when the traffic overload has occurred in the selected access switching processor, setting an originating call approval value at a value resulting from the deduction of a sum of terminating calls, input trunk calls and output trunk calls generated for the predetermined interval from the call approval value setting an input trunk call approval value at a value resulting from the deduction of the sum of terminating calls and output trunk calls generated from the predetermined interval from the call approval value, and setting a terminating and output trunk call approval value at the non-traffic approval value or the traffic approval value;

(d-5) when both the non-traffic approval value and the traffic-approval value should be calculated, setting an originating call approval value at a smaller one of the originating call approval value calculated by use of the traffic approval value and the non-traffic originating call approval value, and setting an input trunk call approval value at a smaller one of the traffic approval value and the non-traffic input trunk call approval value;

(d-6) executing a control for originating calls and input trunk calls by accepting numbers of originating calls and input trunk calls respectively corresponding to the corresponding approval values among calls generated for the predetermined interval in the selected access switching processor while suppressing the remaining calls; and

(d-7) executing a control for terminating and output trunk calls by informing the number translating processor of the terminating and output trunk call approval value by the selected access switching processor and then accepting the number of terminating and output trunk calls corresponding to the terminating and output trunk approval value among calls generated for the predetermined interval in the number translating processor while suppressing the remaining calls.