Security and Privacy in Collaborative eHealth: Expressive and Efficient Attribute-based Access Control with Revocation = Collaborative eHealth 환경에서의 보안 및 개인 정보: 철회 기능을 갖춘 표현적이고 효율적인 속성 기반 접근 제어

다국어 초록 (Multilingual Abstract)

Collaborative ehealth is a paradigm that allows a patient to be treated simultaneously by multiple healthcare service providers through the sharing of the patient’s health data stored in the cloud. However, the use of a third-­party for data storage presents several concerns. In particular, privacy and security concerns have become the biggest challenges in collaborative ehealth systems. Attribute Based Encryption (ABE) which provides fine-grained access control is envisioned to guarantee security and data privacy during health data sharing
in collaborative ehealth systems. In ABE, an access policy is defined and included either in a data user’s secret key or a ciphertext, and decryption is only successful if the access policy is satisfied by an attribute set.

A critical analysis of the related literature reveals that the existing schemes for secure sharing of health data in collaborative ehealth systems can be enhanced in terms of expressiveness, revocation, security, and usability. With regard to expressiveness, the existing schemes rely on access structures that are either monotonic or restrictive and inefficient. With regard to security and attribute/user revocation, the existing ABE schemes insecurely or inefficiently revoke access rights from compromised or obsolete data users. And, regarding
usability, we consider computational efficiency and unboundedness.

Thus, in this thesis, we present two (2) contributions to address the stated limitations for secure and efficient sharing of health data in collaborative ehealth systems.

First, we propose, design and evaluate an expressive access control scheme with revocation for secure and efficient sharing of health data in collaborative ehealth systems. The aim of the proposed scheme is to achieve expressiveness, immediate attribute/user revocation, efficiency, and security. We leverage OBDD for expressiveness and efficiency. The attribute group concept in which each group has a unique key that changes when a group member is revoked of an attribute is used for immediate and efficient attribute/user revocation. For security, we bind decryption keys to user identities. Extensive security and performance analysis of the scheme shows that the proposed access control is expressive, secure, revocable and optimally efficient.

The second contribution is the design and evaluation of CESCR scheme for efficient and secure sharing of health data in collaborative ehealth systems. The CESCR scheme extends the idea of the first scheme specifically for improved usability. The usability improvement is in terms of unboundedness and improved computational efficiency. For unboundedness, we limit the attribute elements in ciphertexts of CESCR scheme to only those associated with group keys. For improved efficiency, the CESCR scheme securely outsources the
computationally demanding attribute operations in encryption and decryption to the cloud with no need for a dummy attribute. A comprehensive security analysis of the CESCR scheme shows that the scheme preserves data confidentiality, and it is resistant to collusion and forward/backward attacks. The performance analysis in terms of storage and computation cost shows that the CESCR scheme is more efficient in comparison with the most related scheme. In general, the CESCR scheme has properties of being expressive, efficient,
unbounded, revocable and secure.

국문 초록 (Abstract)

Collaborative ehealth 는 클라우드에 저장된 환자의 건강 데이터를 공유하여 다수의 헬스케어 제공자가 치료 방식을 제안하는 패러다임을 말한다. 하지만 이 때, 건강 데이터 저장을 위해 서드 파티를 거치면서 여러 문제점이 발생한다. 특히 프라이버시와 보안 문제는 Collaborative ehealth 시스템에서 가장 큰 이슈다. 속성 기반 암호화 (Attribute Based Encryption) 는 Collaborative ehealth 시스템에서 건강 데이터를 공유하는 중 발생하는 프라이버시 문제를 해결하기 위해 제안되는 세밀한 접근 방법이다. 속성 기반 암호화에서 접근 정책은 사용자의 비밀 키나 암호 문 내에 정의 및 포함된다. 또한, 사용자의 속성 집합이 접근 정책을 만족하는 상황에서만 복호화가 가능하다.

관련된 연구를 비판적으로 분석했을 때, Collaborative ehealth 시스템에서 건강 데이터의 안전한 공유를 위해 존재하는 scheme 들은 표현력, 보안, revocation, 유용성 측면에서 향상될 수 있다. 표현력과 관련하여, 기존의 scheme 들은 접근 구조에 의존하기 때문에 단조롭거나 제한적이며, 효율적이지 않다. 보안과 속성/사용자 revocation 과 관련하여, 존재
하는 속성 기반 암호화 scheme 들은 손상되었거나 더 이상 사용하지 않는 데이터 사용자의 접근 권한을 안전하지 않거나 효율적이지 못한 방식으로 revoke 한다. 마지막으로 유용성과 관련하여 본 연구는 계산 효율성과 unboundedness 를 고려한다.

따라서, 본 연구는 앞서 언급한 한계점을 해결하여 Collaborative ehealth 시스템에서의 안전하고 효율적인 건강 데이터 공유를 위한 두가지 contributions 를 제안한다.

첫번째로, Collaborative ehealth 시스템에서 안전하고 효율적인 건강 데이터 공유를 위한 revocation 을 갖춘 표현적인 접근 제어 scheme 을 제안한다. 제안하는 scheme 의 목표는 표현력, 즉각적인 속성/사용자 revocation, 효율성, 보안성이다. 표현력과 효율성을 위해 Ordered Binary Decision Diagram 접근 구조를 활용한다. 속성 그룹 개념을 통해 즉각적
이고 효율적인 속성/사용자를 해지한다. 본 연구는 보안성을 위해서 해독 키를 사용자 identities 와 결합하였다. 광범위한 Scheme 의 보안과 성능 분석은 제안하는 접근 제어가 표현력 및 안전성이 높고, revocable 하며 최적으로 효율적이라는 것을 보여준다.

두번째로, Collaborative ehealth 시스템에서 효율적이고 안전한 건강 데이터 공유를 위해 CESCR scheme 을 디자인하고 평가한 것이다. CESCR scheme 은 유용성을 향상시키기 위해 첫번째 scheme 의 개념을 확장하였다. 유용성 향상은 효율성 개선과 unboundedness 측면에서 이루어진다. Unboundedness 를 위해 본 연구는 그룹 키와 연관이 있는 CESCR scheme 의 암호문 내의 속성 요소를 제한했다. 효율성 개선을 위하여 CESCR scheme 은 계산 비용이 높은 암호화 및 복호화의 연산 과정을 dummy 속성이 필요 없는 클라우드로 안전하게 outsource 하였다. CESCR scheme 의 포괄적인 보안 분석을 통해 scheme 이 데이터 기밀성을 유지하며 collusion 및 순방향/역방향 공격에도 견고하다는 것을 알 수 있다. 또한, 데이터 저장과 계산 비용 측면에서의 성능 분석은 CESCR scheme 이 가장 연관된 scheme 에 비해 우수하다는 것을
보여준다. 일반적으로, CESCR scheme 은 표현력이 높고, 효율적이며, unbounded, revocable, 안전한 속성을 지닌다.

목차 (Table of Contents)

• 1 Introduction 1
• 1.1 Collaborative eHealth 1
• 1.2 Attribute-­based Access Control 3
• 1.3 Motivation and Challenges 6
• 1.4 Research Aim and Objectives 10
• 1.5 Research Methodology 11
• 1.5.1 Literature Review 12
• 1.5.2 Theoretical Work 12
• 1.5.3 Analysis and Evaluation 13
• 1.6 Research Contributions 13
• 1.7 Thesis Organization 16
• 1.8 Publications 16
• 1.8.1 Related Publications 17
• 1.8.2 Other Publications 17
• 2 Literature Review 18
• 2.1 Chapter Introduction 18
• 2.2 Overview of Collaborative eHealth Systems 18
• 2.2.1 Participants 19
• 2.2.2 Security and Privacy Challenges in Collaborative eHealth Systems 20
• 2.2.3 Access Controls for Security and Privacy Provision in Collaborative eHealth Systems 21
• 2.2.4 Requirements Specifications for Health Data Sharing in Collaborative eHealth Systems 23
• 2.3 Attribute-­based Access Control 25
• 2.3.1 ABE Overview 25
• 2.3.2 Key­-Policy Attribute Based Encryption (KP­-ABE) 26
• 2.3.3 Ciphertext-Policy Attribute Based Encryption (CP-ABE) 28
• 2.3.4 Multi­-Authority Attribute Based Encryption (MA­-ABE) 29
• 2.3.5 Hierarchical Attribute Based Encryption (HABE) 31
• 2.3.6 Hidden-­policy Attribute Based Encryption (HP-­ABE) 33
• 2.3.7 Revocable Attribute Based Encryption Schemes 35
• 2.4 Existing Challenges and Solutions for Using ABE in Collaborative eHealth Systems 39
• 2.5 A Way Forward 41
• 2.6 Chapter Summary 44
• 3 Attribute Based Encryption: A Technical Overview 45
• 3.1 Chapter Introduction 45
• 3.2 CP­ABE Model 45
• 3.3 CP­ABE Security Model 46
• 3.4 Technical Preliminaries 47
• 3.4.1 Bilinear Maps 48
• 3.4.2 Security Complexity Assumptions 48
• 3.4.3 Access Structures 49
• 3.5 Ordered Binary Decision Diagram (OBDD) 50
• 3.5.1 OBDD Access Structure 51
• 3.5.2 Satisfying OBDD Access Structure 52
• 3.5.3 OBDD Access Structure Examples 52
• 3.6 The CP­ABE Scheme 56
• 3.6.1 The Construction 56
• 3.6.2 Security Proof 57
• 3.7 Chapter Summary 59
• 4 Collaborative eHealth Security and Privacy: An Expressive Access Control Based on Ordered Binary Decision Diagram (OBDD) Access Structure with Revocation 60
• 4.1 Chapter Introduction 60
• 4.2 The Scheme Contributions 60
• 4.3 Design Preliminaries 61
• 4.3.1 Assumptions 62
• 4.3.2 Threat Model 62
• 4.3.3 Design Requirements 63
• 4.4 System and Security Models 64
• 4.4.1 System Architecture 64
• 4.4.2 System Model 66
• 4.4.3 Security Model 68
• 4.5 The proposed access control scheme construction 69
• 4.5.1 Overview 69
• 4.5.2 The construction 70
• 4.6 Security Analysis 77
• 4.6.1 Data Confidentiality 78
• 4.6.2 Collusion Resistance 81
• 4.6.3 Forward/Backward Secrecy 83
• 4.7 Performance Analysis and Comparisons 84
• 4.7.1 Feature Analysis 84
• 4.7.2 Computational Cost Analysis 84
• 4.7.3 Experimental Results 86
• 4.8 Chapter Summary 92
• 5 CESCR: CP­ABE for Efficient and Secure Sharing of Data in Collaborative eHealth with Revocation 93
• 5.1 Chapter Introduction 93
• 5.2 The Idea 94
• 5.2.1 Overview 94
• 5.2.2 Contributions 96
• 5.3 Design Preliminaries 97
• 5.3.1 Assumptions 97
• 5.3.2 Threat Model 98
• 5.3.3 Design Requirements 99
• 5.4 System and Security Models 100
• 5.4.1 System Architecture 100
• 5.4.2 System Model 102
• 5.4.3 Security Model 104
• 5.5 CESCR scheme construction 106
• 5.6 Security Analysis 112
• 5.6.1 Data Confidentiality 112
• 5.6.2 Collusion Resistance 116
• 5.6.3 Forward/Backward Secrecy 119
• 5.7 Performance Analysis and Comparisons 119
• 5.7.1 Feature and Storage Cost Analysis 119
• 5.7.2 Theoretical Computation Cost Analysis and Comparison 121
• 5.7.3 Experimental Performance Evaluation 123
• 5.8 Chapter Summary 128
• 6 Conclusion and Future Work 129
• 6.1 Conclusion 129
• 6.2 Future Work 132
• References 135
• Appendix A Security Proofs based on Hard Problems 148
• Appendix B Cyclic Groups 149
• B.1 Group 149
• B.2 Finite Group 149
• B.3 Cyclic Group 150