ConXSence: Exploiting Context Information for Security & Privacy Enablers

As the use of smartphones and various mobile applications is increasing, also security and privacy management on mobile devices is gaining importance. This is emphasized by the inherent mobility of smartphones and their capabilities to sense and process contextual information like the device’s location, other devices in proximity, accelerometer readings, and other ambient factors like temperature, humidity, luminosity, noise levels, etc. Such sensor data enable to construct a very detailed picture about the behavior and social relations of the device’s use.

Contextual adaptation

Contextual adaptation can also be utilised in more intelligent enforcement of security features. For example by sensing and profiling the mobile device’s context one can implement an adaptive device lock that enforces a long locking time-out in contexts where the risk of device theft or misuse is low, while locking the device quickly in more risky environments like crowded public places.

Hence, there is a need to control the use and release of contextual information collected with the device’s sensors in order to encounter threats arising from applications like sensory malware misusing the sensors of the device to obtain sensitive information about the user. On the other hand we want to utilize the contextual information available in the device’s ambient context in order to enable intelligent context-aware access control.

Context-Aware Access Control

While the idea of context-aware access control is not new, the currently proposed solutions rely on pre-configured access control policies that are either specified by the user or some other party (e.g., application developer, device vendor, network administrator, etc.). The drawback with user-specified policies is that the amount of work required to set up and maintain a comprehensive set of context-dependent policies is significant. It is not likely that average users of mobile devices will be willing to spend such a significant effort in maintaining their policy set. On the other hand, it is also not a trivial task to create a coherent set of policies that correspond to the intentions of the user. It is questionable, whether regular users have the required expertise to set up their policies correctly, resulting thus in erroneous policy sets. A quick remedy could be to resort to default sets of policies defined by expert users like systems administrators, but these can’t take the personal security and privacy preferences of individual users adequately into account.

Context Profiling

However, recently, researchers have started to explore how to improve the usability of context-aware access control by using context information more intensively, e.g., by estimating based on contextual information obtained through sensors the security level of a particular situation [2], or, evaluating the likelihood that the user of the mobile device indeed is its authentic owner [3].

This project investigates therefore methods for using sensed context information for specifying and maintaining context-dependent security and privacy policies [1] and establishing security associations between the users’ devices and other users in a usable and intuitive way, while providing sensible policy settings reflecting the true privacy needs of the user. By monitoring and profiling the user’s behavior and context it is possible to automatically identify relevant contexts, e.g., the most frequently visited places or sets of frequently encountered peer devices. By limited user feedback and monitoring of the user’s behavior we learn appropriate security settings applicable in specific contexts and thus adapt the enforcement behavior of the device to reflect user needs and preferences.

On the other hand, context data themselves can be useful in creating security associations between previously unknown devices. In this project we examine therefore also, how ambient context data can be used in mutual key agreement and key management between devices frequenting similar contexts without the need for explicit manual pairing.


To show the feasibility of our scheme we implemented ConXSense for the Android operating system and integrated it with the Android Security Modules Framework [4] to enable Context-Aware Access Control on applications, sensors and privacy-sensitive resources on Smartphones and Tablets.


[1] Markus Miettinen, Stephan Heuser, Wiebke Kronz, Ahmad-Reza Sadeghi, N. Asokan, “ConXsense – Context Sensing for Adaptive Usable Access Control”, in Proceedings of ASIACCS 2014 [Best Paper Award]

[2] A. Gupta, M. Miettinen, N. Asokan, and M. Nagy, “Intuitive security policy configuration in mobile devices using context profiling”, in SocialCom/PASSAT. IEEE, 2012, pp. 471–480.

[3] O. Riva, C. Qin, K. Strauss, and D. Lymberopoulos, “Progressive authentication: deciding when to authenticate on mobile phones”, in Proceedings of the 21st USENIX Security Symposium, 2012.

[4] S. Heuser, A. Nadkarni, W. Enck and A.-R. Sadeghi, “ASM: A Programmable Interface for Extending Android Security”, in Proceedings of the 23rd USENIX Security Symposium, 2014