Test Automation Framework for Embedded Devices on Cloud : TAF-EDC
- 주제(키워드) Cloud System , Reliability , Test Automation System
- 발행기관 고려대학교 대학원
- 지도교수 정창성
- 발행년도 2021
- 학위수여년월 2021. 8
- 학위구분 박사
- 학과 대학원 전기전자전파공학과
- 세부전공 전자전기컴퓨터공학전공
- 원문페이지 109 p
- UCI I804:11009-000000252075
- DOI 10.23186/korea.000000252075.11009.0001223
- 본문언어 영어
초록/요약
Embedded systems which consisting of an electronic control system with hardware and software to perform specific functions have the characteristic of high coupling and dependency between components of heterogeneous layers in system, which makes it difficult to track the location and cause of faults by testing each component independently. Despite these worst conditions of verification, even reliability quality verification, one of the quality characteristics of embedded systems, is verified with typical test strategies, test cases, and insufficient sample size depending on the limitations of verification cost and development schedule, etc. Since verification considering the quality characteristics of such embedded system is not sufficiently, there is a high possibility that various reliability and functional quality problems may occur after shipment. Hence, to address this gap, we shall present a Test Automation Framework for Embedded Devices on Cloud (TAF-EDC) that can automatically perform various functional and reliability evaluations by exploiting the cloud concept to make the most of distributed and shared embedded devices. Next, we shall present a new innovative Adaptive Sample scale for Regulation Test (ASRT) which can derive the most appropriate sample size for functionality and reliability test of embedded devices to establish a secure test coverage rate by solving the dispersion issues on massive, distributed devices. Finally, the new Mass Sample for Regulation Test (MSRT) which we propose extracts test case and determine the test ordering based on the Markov process to increase the number of times that reliability issues in embedded systems were detected, signifying its effectiveness in identifying problems. In this paper, we address how many pre-detections of reliability problems that occurred with smartphones and whether the results detected are statistically significant through the proposed reliability module of TAF-EDC targeting Android smartphone models, which is considered the most common case including numbers of embedded components.
more목차
TABLE OF CONTENTS
ABSTRACT i
LIST OF FIGURES v
LIST OF TABLES v
Chapter 1. Introduction 1
Chapter 2. Related works 4
2.1 Cloud computing platform 4
2.1.1 Overview 4
2.1.2 Classification 9
2.1.3 Testing platform on cloud 11
2.2 Consideration of sample size 13
2.2.1 Operation characteristics (OC) curves 14
2.2.2 Statistical sample size determining 24
2.2.3 Markov Processes 27
2.3 Embedded system 31
2.3.1 Structure 31
2.3.2 Embedded Software 32
2.4 Software product quality model 38
2.4.1 Overview 38
2.4.2 Software validation flow 39
2.4.3 Quality characteristics in ISO/IEC 9126 41
Chapter 3. TAF-EDC System 45
3.1 Overview 45
3.2 Architecture 45
3.3 Test Framework Manager (TFM) 48
3.4 Sample Device Handler (SDH) 54
3.5 Operation 58
Chapter 4. Reliability test method 67
4.1 Adaptive Sample scale for Reliability Test (ASRT) 67
4.1.1 Structural improvement using cloud concept 67
4.1.2 Statistical prediction of ASRT 71
4.2 Mass Samples Reliability Test Process (MSRTP) 72
Chapter 5. Performance evaluation 76
5.1 Adaptive Sample Scale for Reliability Test (ASRT) 76
5.1.1 ASRT application effect 77
5.1.2 Estimation of test samples using MiniTab 79
5.2 Mass Samples Reliability Test (MSRT) 86
5.2.1 MSRT application 89
5.2.2 MSRT application effect 90
Chapter 6. Conclusion 93
Reference 95
