Markus Endler

An Architecture for Distributed High Performance Video Processing in the Cloud

Video processing applications are notably data intense, time, and resource consuming. Upfront infrastructure investment is usually high, specially when dealing with applications where time-to- market is a crucial requirement, e.g., breaking news and journalism. Such infrastructures are often inefficient, because due to demand variations, resources may end up idle a good portion of the time. In this paper, we propose the Split&Merge architecture for high performance video processing, a generalization of the MapReduce paradigm that rationalizes the use of resources by exploring on demand computing. To illustrate the approach, we discuss an implementation of the Split&Merge architecture, that reduces video encoding times to fixed duration, independently of the input size of the video file, by using dynamic resource provisioning in the Cloud.

Pervasive social context: Taxonomy and survey

As pervasive computing meets social networks, there is a fast growing research field called pervasive social computing. Applications in this area exploit the richness of information arising out of people using sensor-equipped pervasive devices in their everyday life combined with intense use of different social networking services. We call this set of information pervasive social context. We provide a taxonomy to classify pervasive social context along the dimensions space, time, people, and information source (STiPI) as well as commenting on the type and reason for creating such context. A survey of recent research shows the applicability and usefulness of the taxonomy in classifying and assessing applications and systems in the area of pervasive social computing. Finally, we present some research challenges in this area and illustrate how they affect the systems being surveyed.

Middleware: Context Management in Heterogeneous, Evolving Ubiquitous Environments

Mobile computing environments are characterized by heterogeneity - systems consisting of different device types, operating systems, network interfaces, and communication protocols. Such heterogeneity calls for middleware that can adapt to different execution contexts, hide heterogeneity from applications, and transparently and dynamically switch between network and sensor technologies. Additionally, middleware for context-aware systems must keep a context model (a model of their environment), taking into account several aspects of the environment. The more complex and heterogeneous an execution environment is, the more complicated its underlying context model. Moreover, because systems can evolve, context management must also support model evolution without restarting, reconfiguring, or redeploying applications and services. We describe a context management middleware that can efficiently handle context despite the execution environments heterogeneity and evolution. It uses context meta-information to improve a context-aware systems overall performance

A Privacy Service for Context-aware Mobile Computing

Privacy issues related to the access of context information are becoming increasingly important as we move toward ubiquitous and mobile computing environments. In this article, we describe the design and implementation of a privacy service, called Context Privacy Service (CoPS), to control how, when and to whom disclose a user’s context information. Based on the results of an end-user survey and experience reported by other research groups, we identified the main service requirements and designed CoPS aiming flexibility, generality, simplicity and fine-grained privacy control. CoPS is an optional service of our context-provisioning middleware MoCA and allows users of context- and location-aware applications to define and manage their privacy policies regarding disclosure of their context information. The main features supported by CoPS are group-based access control, pessimistic and optimistic approaches for access control, hierarchical privacy rules, mixed-initiative interaction, and rule specificity analysis.

A DDS-based middleware for scalable tracking, communication and collaboration of mobile nodes

Applications such as transportation management and logistics, emergency response, environmental monitoring and mobile workforce management employ mobile networks as a means of enabling communication and coordination among a possibly very large set of mobile nodes. The majority of those systems may thus require real-time tracking of the nodes and interaction with all participant nodes as well as a means of adaptability in a very dynamic scenario. In this paper, we present a middleware communication service based on the OMG DDS standard that supports on-line tracking and unicast, groupcast and broadcast with several thousand mobile nodes. We then show a Fleet Tracking and Management application built using or middleware, and present the performance results in LAN and WAN settings to evaluate our middleware in terms of scalability and robustness.

DICHOTOMY: A Resource Discovery and Scheduling Protocol for Multihop Ad hoc Mobile Grids

There has been a good amount of research over the past few years on the integration of mobile devices into computational grids. Little of it, however, has addressed the more challenging issue of dynamically establishing spontaneous, multihop ad hoc mobile grids, and then only in preliminary form. In such grids, resource discovery and scheduling have been traditionally approached separately in the literature. In our view, this is an inappropriate dichotomy as regards multihop ad hoc mobile grids. In this paper, we propose a resource discovery and scheduling protocol for mobility (DICHOTOMY), which allows resource provisioning to be scheduled among the most resourceful nodes in the mobile grid, while mitigating the overhead of discovery messages exchanged among them. Our experimental results in a testbed and in a simulation platform show that the protocol provides efficient load balancing between nodes under an increasing number of requests, and scales well with respect to an increasing number of nodes in comparison with traditional broadcast-based solutions for service discovery.

Defining Situated Social Context for pervasive social computing

There is a multitude of systems in the area of mobile and pervasive social computing providing and using social context for pervasive interaction between persons. Currently, it is hard to distinguish the different approaches as there is no appropriate taxonomy available yet. In this paper we introduce the STIP taxonomy to define social context in the area of social computing and further propose to use the term Situated Social Context for a more focused subset of social context for pervasive social computing. We elaborate the taxonomy with a small survey of mobile social networking systems and discuss future research challenges.

Evaluating composite events using shared trees

Distributed and concurrent systems tend to generate large numbers of events which require further processing, such as filtering and establishment of co-relations, to become useful for human or automated managers. The work focuses on the detection of composite events defined through event expressions involving primitive events and some operators: Boolean conjunction and disjunction; sequence, repetition, and absolute, periodic and relative timer events. A concrete result of this work was the design and implementation of a general purpose event processing service (EPS) which may be used by monitoring applications to be informed about the occurrence of primitive and composite events. Composite events are commonly represented as trees where the leaf nodes represent primitive event types and the intermediate nodes represent any of the supported event operators. The main contribution of this work is the description of a method to process composite events that share common sub-expressions with other composite events. The EPS was implemented in Java within the framework of the Sampa project.

Evolutionary and efficient context management in heterogeneous environments

Mobile computing and pervasive environments are mainly characterized by heterogeneity of devices, with different capabilities, resources, operating systems and applications. In a realistic scenario for context-aware computing, middleware should be deployable in the whole distributed system, despite devices resource limitations, and the developer should be able to evolve the context model when new context-aware applications or context providers are introduced. This paper discusses how context modeling and design of middleware architecture can impact on the efficiency of provision, distribution and access of context information in heterogeneous environments. This paper describes a middleware architecture and design strategies in order to address such requirements.

Support for Context-Aware Collaboration

This paper describes a middleware architecture with its location inference service (LIS), and an application for context-aware mobile collaboration which is based on this architecture. The architecture, named Mobile Collaboration Architecture - MoCA comprises client and server APIs, a set of core services for registering applications, monitoring and inferring the execution context of mobile devices, in particular their location. This architecture is suited for the development of new kinds of collaborative applications in which the context information (connectivity, location) plays a central role in defining both the group of collaborators, and the communication mode.