Mapa mental
ContexIoT
Conteúdo do mapa
ContexIoT
- 1. Appified IoT: 3rd-party develoepers 2. Permission model are flawed--> in need of a new access control model 3. ContexIoT: context-based permission system: (1). fine-grained context identification for sensitive actions (2). runtime prompts for users. 4. Context definition in ContexIoT is at inter-procedure level, and data flow level. 5. Prototyped on Samsung SmartThings.
Security Analysis of Emerging Smart Home Applications (IEEE S&P 2016)
- Security design flaws in the permission model of IoT platforms.
- Therefore, a new access control model is needed in these appified IoT platforms
- Access control mechanism of Recent appified platform (such as smartphone) are coarse-grained, insufficient and undemanding
- [19] SoK: Lessons Learned From Android Security Research For Appified Software Platforms (IEEE S&P 2016)
- User-driven access control: Rethinking permission granting in modern operating systems (IEEE S&P 2012)
- [64] Android permissions remystified: A field study on contextual integrity (USENIX Security 2015)
- Aurasium: Practical Policy Enforcement for Android Applications (USENIX Security 2012)
- From numerous studies on Android and iOS permission systems, a key design flaw is that they either require users to make uninformed decision at install time [1] or prompt users at runtime when an app requests any of a handful of resources, without providing essential contextual information [2], [8]. These studies conclude that it is highly desirable to put the user in context when making permission granting decisions at runtime. This helps ensure a property known as “contextual integrity” defined by Nissenbaum [52] with which “information flows according to contextual norms,” and it is advocated as the desired norm for future permission system design of the smartphone platform [64], [68], [19].
- [52] Privacy as contextual integrity (2004)
- [68] Appcontext: Differentiating malicious and benign mobile app behaviors using context. (IEEE/ACM 2015)
- [19] SoK: Lessons Learned From Android Security Research For Appified Software Platforms (IEEE S&P 2016)
- As discussed in previous attempts to support contextual integrity in smartphone, providing contextual integrity in appified system is challenging because
- Availability of context: sensitive permission requests occur when the user is not interacting with the requesting app [64]
- Frequency of prompt: too frequent is annoying
- A permission is granted by the user to a functionality only under a particular usage context
- Usage Context of an app functionality=execution flow at runtime (trigger + callgraph) + data flowing through that path
- Practical trigger-action programming in the smart home (SIGCHI 2014)
- 33.21.50.67.19.64..56.54///usability34.64.44
- Usage Context of an app functionality=execution flow at runtime (trigger + callgraph) + data flowing through that path
- [68] Appcontext: Differentiating malicious and benign mobile app behaviors using context. (IEEE/ACM 2015)
- [52] Privacy as contextual integrity (2004)
- [64] Android permissions remystified: A field study on contextual integrity (USENIX Security 2015)
- User-driven access control: Rethinking permission granting in modern operating systems (IEEE S&P 2012)
- [19] SoK: Lessons Learned From Android Security Research For Appified Software Platforms (IEEE S&P 2016)
How do they patch Samsung SmartThings
- To support existing SmartApps without changing the close-source SmartThings cloud backend, we developed an app patching machanism that convert unmodified SmartApps to ContexIoT-compatible SmartApps. The patching process separates the execution flow of a sensitive action in the original SmartApp into 2 steps: (1). collect the context info before the action is executed.
Extensive study of possible attacks on appifiede IoT platforms
Taxonomized IoT attack app dataset (based on lifecycle and adversary techniques
Informações
- Descrição
- Mapa Mental sobre ContexIoT, criado por David Liu em 31-03-2017.
- Criado por
- David Liu · mar 2017
- Atualizado
- 1 abr 2017
- Visualizações
- 3
- Avaliação