The popularity of the Android System in combination with the lax market approval process may attract the injection of malicious applications (apps) into the market. Android features a permission system allowing a user to review the permissions an app requests and grant or deny access to resources prior to installation. This system conveys a level of trust due to the fact that an app only has access to resources granted by the stated permissions. Thereby, not only the meaning of single permissions, but especially their combination plays an important role for understanding the possible implications. In this paper we present a method that circumvents the permission system by spreading permissions over two or more apps that communicate with each other via arbitrary communication channels. We discuss relevant details of the Android system, describe the permission spreading process, possible implications and countermeasures. Furthermore, we present three apps that demonstrate the problem and a possible detection method.

The increasing market share of the Android platform is partly caused by a growing number of applications (apps) available on the Android market: by now (January 2011) roughly 200.000. This popularity in combination with the lax market approval process attracts the injection of malicious apps into the market. Android features a fine-grained permission system allowing the user to review the permissions an app requests and grant or deny access to resources prior to installation. In this paper, we extract these security permissions along other metadata of 130.211 apps and apply a new analysis method called Activation Patterns. Thereby, we are able to gain a new understanding of the apps through extracting knowledge about security permissions, their relations and possible anomalies, executing semantic search queries, finding relations between the description and the employed security permissions, or identifying clusters of similar apps. The paper describes the employed method and highlights its benefits in several analysis examples – e.g. screening the market for possible malicious apps that should be further investigated.

The security of virtualization platforms can be improved by applying trusted computing mechanisms such as enforcing the integrity of the hypervisor. In this paper we build on a recently proposed platform that extends this trust on to applications and services. We describe a process that covers the fully integrity-enforcing life-cycle of a trusted virtual application. Our architecture allows applications the safe transition between trusted states, even in case of updates of the hypervisor. We also detail the technical realization in our prototype implementation.

Modern Trusted Computing platforms offer the basic hardware building blocks to allow effective enforcement of software integrity. In this paper we present a practical software system architecture which uses Intel’s late launch mechanism to boot a known-good configuration. We restrict the access to data and execution of services to trusted platform configurations, enforcing the integrity of contained applications as specified by the platform operator. Further, we also describe a set of operational procedures to allow flexible and dynamic configuration management. We present our prototype implementation which integrates well with established Linux distributions.