Generating Test cases for Testing Embedded Systems using Combinatorial Techniques and Neural Networks based Learning Model
Sasi Bhanu Jammalamadaka · International Journal of Emerging Trends in Engineering Research · 2019
Embedded Systems must be bug-free.Comprehensive Testing of embedded systems is required to eliminate the risk and reduce Time and also to improve the performance of the embedded systems.Tools have been in existence for undertaking different kinds of Testing that include integraTion and system Testing.Embedded systems must be tested considering different perspecTives that include input, output, mulTi-input, mulTi-output, and input-output relaTionships.No tool is in existence for Testing the embedded systems, both the information and output domain.There is a need to learn a model represenTing the embedded systems, and the output generated out of the model must tally with the expected production to guarantee that the embedded system is tested comprehensively considering the input and output domain.In this paper, a method has been presented that learns the embedded system through building the Neural network and the correctness or exactness of the model through comparing the outputs generated through Neural network with the outputs collected through manual Testing.The paths exisTing in the firmware that does not yield the expected output are determined and listed, so that code relaTing to those paths is corrected.The Testing method proposed in this paper covers the Input-Output perspecTive of the embedded system