Optimum functional test method to determine the quality of a software system embedded in a large electronic system

• US 5,500,941 A
• Filed: 07/06/1994
• Issued: 03/19/1996
• Est. Priority Date: 07/06/1994
• Status: Expired due to Term

• Current Assignee: Ericsson SA (Telefonaktiebolaget LM Ericsson)
• Original Assignee: Ericsson SA (Telefonaktiebolaget LM Ericsson)
• Inventors: Gil, Luis
• Primary Examiner(s): Beausoliel, Jr., Robert W.
• Assistant Examiner(s): Palys, Joseph E.
• Application Number:
  US08/271,422
• Time in Patent Office: 622 Days
• Field of Search: 395/575, 395/183.14, 395/183.01, 395/183.13, 395/185.1, 371/19, 371/23
• US Class Current: 714/38.1
• CPC Class Codes: G06F 11/3684 for test design, e.g. generating new test cases

Abstract

A method of performing software validation testing on large electronic systems applies Usage Concepts specially modelled as Markov chains to forrealize the expected use of the system, to define the system behavior, and to introduce statistical measurements. The functional requirements of the software system are defined in a Markov chain by identifying a plurality of states through which the software system transitions when performing the functional requirements. Stimuli and responses which cause the software system to transition from one state to another are identified, and the probability of each state transition is calculated. Usage cases beginning at an invocation state and ending at a termination state of the software system are defined and listed in order of descending probability of occurrence in a Usage Profile. The Usage Profile, stimuli, and responses are automatically compiled into an automated test program, and associated test equipment is automatically programmed to generate a set of test cases. The set of test cases are then executed on the software system. The end of the set of test cases is determined, and the method analyzes and reports the software system quality.

Claims

1. An electronic media for a stored program for electronic systems wherein said stored program performs the method of performing validation testing to determine the quality of a software system embedded in a large electronic system having associated test equipment, said method comprising the steps of:

defining a set of functional requirements for said software system;

constructing a Markov chain, said constructing step further comprising the steps of,

identifying a plurality of states through which said software system transitions when performing said functional requirements;

defining a plurality of stimuli which cause said software system to transition from one of said plurality of states to another;

identifying a response, comprising at least one state transition, which occurs in response to each of said plurality of stimuli for each of said plurality of states; and

calculating the probability of occurrence of each of said state transitions;

identifying a plurality of usage cases, each of said usage cases comprising a set of state transitions beginning at an invocation state and ending at a termination state of said software system;

building a Usage Model for said software system, said Usage Model comprising a plurality of usage cases;

calculating the probability of occurrence of each of said usage cases;

building a Usage Profile for said software system by listing said usage cases in order of descending probability of occurrence;

automatically compiling said Usage Profile, said stimuli, and said responses into an automated test program;

providing a stored program media containing the programming for said associated test equipment with said automated test program to generate a set of test cases;

executing said set of test cases on said software system;

determining the end of said set of test cases; and

analyzing and reporting said software system quality.

2. An electronic media of claim 1 wherein said program'"'"'s said step of building a Usage Model further includes the steps of:

translating characteristics of said software system into parameters of said Usage Model; and

statistically analyzing said Usage Model parameters to determine the extent to which said parameters exhibit characteristics of said software system.

3. An electronic media of claim 2 wherein said program'"'"'s step of building a Usage Model includes iteratively repeating the steps of translating characteristics of said software system into parameters of said Usage Model and statistically analyzing said Usage Model parameters, until a percentage of said Usage Model parameters exhibiting characteristics of said software system exceeds a predetermined level.

4. An electronic media of claim 3 wherein said program'"'"'s said step of building a Usage Model for said software system includes automatically generating said usage cases using a random number generator and a high level software language.

5. An electronic media of claim 1 wherein said program'"'"'s said step of building a Usage Profile for said software system includes loading said Usage Model into an analyzer to explore all possible usage cases transitioning between two states in said Usage Model.

6. An electronic media of claim 5 wherein said program'"'"'s said step of building a Usage Profile for said software system includes loading said Usage Model into an analyzer in the form of a probability transition matrix, a state-stimulus matrix, and a state-expected response matrix.

7. An electronic media of claim 6 wherein said program'"'"'s said step of building a Usage Profile for said software system includes the steps of:

eliminating usage cases from said Usage Profile if the total probability of a given usage case is less than a first predetermined threshold value;

eliminating usage cases from said Usage Profile if the probability of a particular single transition in a given usage case is less than a second predetermined threshold value; and

eliminating usage cases from said Usage Profile if the number of state transitions in a given usage case exceeds a predetermined number.

8. An electronic media of claim 7 wherein said program'"'"'s said step of building a Usage Profile for said software system includes setting an upper limit on the number of usage cases in the Usage Profile, said step further comprising the steps of:

causing at least two runs of said analyzer;

reducing the second threshold value by a predetermined amount for each run of said analyzer;

generating a file which contains the result of each run of said analyzer; and

stopping said analyzer runs when the probabilities of all the usage cases with the same number of transitions is less than one-fifth the life of the software system.

9. An electronic media of claim 8 wherein said program'"'"'s said step of building a Usage Profile for said software system includes the steps of:

increasing said first and second threshold values by a predetermined amount for each run of said analyzer;

generating a file which contains the result of each run of said analyzer;

stopping said analyzer runs when a run results in zero usage cases; and

creating a final list of usage cases by appending all of the files generated by the runs of the analyzer and deleting repeated usage cases.

10. An electronic media of claim 9 wherein said program'"'"'s said step of building a Usage Profile for said software system includes the steps of:

sorting the usage cases in said final list of usage cases by probability of occurrence; and

numbering the usage cases in order of decreasing probability of occurrence.

11. An electronic media of claim 10 wherein said program'"'"'s said step of automatically compiling said Usage Profile, said stimuli, and said responses into an automated test program includes compiling said Usage Profile, said stimuli, and said responses into a single file formatted to control said associated test equipment for said large electronic system.

12. An electronic media of claim 11 wherein said program'"'"'s said step of executing said set of test cases includes executing said set in order of decreasing probability of occurrence.

13. An electronic media of claim 12 wherein said program'"'"'s said step of determining the end of said set of test cases includes the steps of:

calculating a failure probability for the software system following execution of-each test case; and

stopping the execution of said test set when the failure probability of said software system reaches a predetermined level.

14. An electronic media of claim 12 wherein said program'"'"'s said step of determining the end of said set of test cases includes stopping the execution of said test set following the execution of a predetermined number of test cases.

15. An electronic media of claim 14 wherein said program'"'"'s said step of analyzing and reporting said software system quality includes statistically determining a failure probability for the software system.