Ensuring the software quality of mobile applications with respect to performance, robustness, energy consumption, security and privacy is an important problem for a growing researcher and developer community. In this paper, we present Dynalize, a Platform-as-a-Service cloud for the dynamic analysis of mobile applications. It allows researchers and developers to investigate mobile applications at runtime in a virtual device cloud and to publish the performed analyses as web services. In contrast to existing approaches, it makes use of container virtualization on top of Infrastructure-as-a-Service instances, enabling dynamic provisioning and fast deployment of dynamic analyses. A custom container layout and a novel storage solution on the virtual server layer ensures cost- and runtime-efficient large-scale analyses of thousands of apps. The applicability of Dynalize is demonstrated by a security analysis of about 6,000 Android applications. Experiments on container startup, virtual device to container throughput and different storage back ends show the feasibility of the proposed approach.

Improving the energy efficiency of software running in a data center is a challenging task. Several application-specific techniques, such as energy-aware heuristics, controlled approximation and energy-conserving I/O, have been proposed to tackle this problem. In this paper, we introduce data sparsing with artifacts, a novel approach to increase the energy efficiency of applications that are robust to input variations, such as speech and image processing. Data sparsing with artifacts is aimed at reducing the processing times and thus the energy efficiency of such applications while preserving the quality of the results by replacing a random subset of the original data with application-specific artifacts. In contrast to previous work, the proposed approach introduces artifacts at the data layer, without application layer modifications and with general purpose hardware. Data sparsing with artifacts has been integrated into a prototypical file system in userspace (FUSE) and the Hadoop Distributed File System (HDFS). Experiments with MapReduce-based face detection, face recognition and speech recognition algorithms show promising energy savings of up to 10 % with moderate accuracy losses for different data sparsing rates and artifacts.