PhD Forum

PhD Student Abstracts

The following topics will be presented during the NetSys 2015 PhD Forum poster session.

  • PDFCombining Opportunistic and Information Centric Networks in Real World ApplicationsWith the ever increasing number and power of computing devices around us, we envision a world where we are surrounded by information. Opportunistic Networking could play a role in this scenario and we explore approaches to improve Opportunistic Networks by exploiting the new paradigm of Information Centric Networking.
    Edo Monticelli (University of Goettingen)
  • PDFContext Prediction for Parallel Task Distribution in Highly Dynamic Mobile NetworksMobile devices such as tablets, smartphones or wearables are not only increasing in number but also in processing power. Hence, such devices can be connected to a mobile grid to perform parallel calculation of tasks. However, mobile grids are highly dynamic so that parallel computation can not be performed with sufficient reliability.
    In this paper we propose an approach to build a system that uses context information of mobile devices in the grid to build a context history. This data is then used to predict the availability and status of a device at a given point in time. Our approach is based on the assumption that context-aware node prediction will enable optimization of dynamic task management in a mobile crowd.

    Richard Stück (Lübeck University)
  • PDFContext-Aware and Class-Based Broadcasting in VANETsDedicated Short Range Communication (DSRC), i.e., the use of wireless communication for message dissemination, is a major building block in Inter-Vehicle Communication (IVC). We investigate broadcasting as the main communication primitive to share information among vehicles, or between vehicles and infrastructure. In recent years dozens of broadcast protocols have been proposed, including protocols for one-hop broadcasting (beaconing) of status information and protocols for geo-casting. The main commonalities of all these solutions are that they are designed only for one specific application and cannot co-exist, nor a single broadcast solution can meet the demands of all applications. This PhD thesis focuses on a holistic Network layer to support broadcast based applications.
    Florian Klingler (University of Paderborn)
  • PDFContext-Aware Delivery of Interaction Tutorials in Ambient SpacesOver the past years, the amount as well as the provided functionality of (smart) devices raised, with increasing tendency. In terms of ubiquitous computing and smart home solutions, systems get more embedded into ambient spaces. Recent studies showed, that users are often overwhelmed even in interacting with everday objects, which can be applied to smart objects and gesture control as well. In addition to an insufficent design a lack of documentation is responsible for this.
    In this paper we present an approach to create a framework based on formal descriptions of interaction, space and objects to allow a context-aware delivery of interaction tutorials, based on the users needs and preferences.

    Daniel Burmeister (University of Lübeck)
  • PDFDeveloping Serious Games in Pervasive EnvironmentsWhile students bring a high variety of mobile devices into lecture rooms, they are until today not utilized for an enriched learning experience. With the help of a pervasive middleware for serious games, it is possible to incorporate these student devices into the learning process together with additional devices, such as public displays and data glasses. The middleware is responsible for handling the game state and communicating with the connected devices. The pervasive game middleware is accompanied by developer tools for the creation of serious games. The developer tools are an easy to use Visual Programming Environment and enable people without a technology background to create serious games within pervasive lecture rooms. This leads to a more interactive and playful learning experience for future students.
    Jens Naber (University of Mannheim)
  • PDFDistributed Secure Monitoring in Large Scale Overlay NetworksControlling the Quality of Service of peer-to-peer networks may open doors to new quality-centric overlays and application fields, such as peer-to-peer-based online social networks, telemedicine or robot coordination. To reach this aim, a sophisticated, timely and broad view on the current status of the network, the nodes and the content in the network is needed, in order to come to optimized decisions in the various protocols operating in the peer-to-peer system. The main aim of our research topic is to research monitoring mechanisms which provide sophisticated information on the network’s status, the nodes and the content. Our mechanisms are evaluated within large-scale simulations as well as in peer-to-peer applications tested with real users. For the latter, we developed a Web-based peer-to-peer framework which does not require the users to install any software installation to join. It supports a simple secure online social network and is capable to support large-scale networks. This research statement contains aims and challenges of our work.
    Andreas Disterhöft (University of Düsseldorf)
  • PDFEnhancing Greedy Embedding for Dynamic Scenarios in DarknetsAnonymous communication and censorship resistant data storage through networks of fixed trusted connections become more important to overcome trust problems of centralized approaches. Currently, most discussed darknet approaches, such as the darknet mode of Freenet, base on heuristic routing and embedding methods, which are inefficient and vulnerable to certain attacks. A newly presented greedy embedding could overcome those problems. It can provide a guaranteed path of reasonable length for a connection between arbitrary peers, yet it has only been shown to work in comparatively static environments. To make this approach feasible for more realistic dynamic scenarios with node churn and to overcome known problems, it has to be developed further. Its construction phase is a vulnerable point, in a dynamic scenario this becomes important to solve. Also, it still has problems distributing load because some peers would have to handle more connections going through than others depending on the embedding’s structure. The overall goal is to find out, if the greedy embedding approach can be advanced such that it is more efficient than a comparable heuristic approach and still does not raise problems regarding anonymity.
    Jan-Michael Heller (TU Darmstadt)
  • PDFHeterogeneous Vehicular NetworksCar-to-car and car-to-infrastructure communication enables a lot of different application scenarios. The most popular way of using WLAN communication between cars might not be good enough for every use case. If as an alternative cellular networks are used, capacity problems arise. This is why so called heterogeneous vehicular networks attract more and more attention. Cars that make use of such networks are usually
    equipped with WLAN and cellular devices. Such a combination is able to overcome shortcomings of one technology by using the other one. Still, the tools used to evaluate such algorithms are not sufficient. I want to fix this by adding heterogeneous vehicular networking functionality to existing tools. Additionally, I want to investigate how multiple heterogeneous vehicular networking algorithms work when used together. A possible outcome could be more general concepts, with a focus on clustering, to enable multiple algorithms running beside each other.

    Florian Hagenauer (University of Paderborn)
  • PDFHybrid IEEE 802.15.4 Network EmulationThis research statement describes the author’s PhD project, which targets the evaluation of sensor network designs in realistic environments. Network emulation includes a variety of methods and strategies for the evaluation of system designs for networked systems in safe testing environments. The author’s interest is to bridge the gap between generic simulation environments and real sensor network deployments for IEEE 802.15.4 Wireless Sensor Networks (WSNs) through a combination of network simulators like OMNeT++ and RoSeNet, a network emulation platform.
    Sebastian Böhm (Brandenburg University of Technology Cottbus–Senftenberg)
  • PDFLocation Privacy for Movement TrajectoriesWith widespread adoption of location-enabled devices in the past decade such as smart-phones, location-based services are penetrating fast into our daily lives. While these services provide highly useful information based on user location, their usage raises serious privacy concerns. This proposal aims to address the problem of location privacy in continuous sharing of location data with third parties who are trusted by users at varying levels.
    Zohaib Riaz (University of Stuttgart)
  • PDFNetwork Monitoring for Todays InternetThe task of network monitoring gives operators the ability to analyze their traffic in order to keep their communication network up and running. Monitoring appliances have to adapt continuously to the changes in how the Internet is used. The increase in functionality, usability and secrecy necessitates revisions in monitoring mechanisms. The challenges are manifold: The shift of transport services to the application layer or the rising usage of encryption makes monitoring evermore complex. The solutions currently available are addressing largely individual issues only and are mostly closed source. This PhD thesis will tackle monitoring issues related to high bandwidth links and encryption as well as problems that came up due to the changed usage of HTTP.
    Felix Erlacher (University of Innsbruck)
  • PDFPrivacy in Smart GridsCurrent smart grid systems and algorithms use a lot of private data to perform energy management. This paper proposes an approach to achieve a universally applicable system for privacy in distributed energy management systems.
    Daniel Brettschneider (University of Applied Sciences Osnabrück)
  • PDFRobust Execution of Processes and Management of Heterogeneous Knowledge in Collective Adaptive SystemsThe recent advances in pervasive and mobile computing have paved the way for Collective Adaptive Systems (CAS). The vision of such systems is to enable a network of heterogeneous computing nodes, ranging from highly mobile smart phones to cloud servers, to collaborate in order to benefit from synergistic effects. The core of CAS is the robust and collaborative execution of processes across system boundaries, combined with continuous learning and adaptation. Thus, robust process execution and scalable management of learned knowledge are the key success factors of such systems. This proposal addresses methods to increase the robustness of process execution in such systems and methods for efficient and scalable management of distributed knowledge.
    Thomas Bach (University of Stuttgart)
  • PDFSemantic-based Management of Federated Infrastructures for Future Internet ExperimentationIn the context of distributed computing various strategies have been developed to share heterogeneous resources across multiple administrative domains. An important challenge that arises in this context is the exchange of information about the provided resources and their dependencies. In the considered field of application of federated Future Internet testbeds, existing work rest upon different interfaces and schematic data models. This aggravates the discovery and management of heterogeneous resources between autonomous facilities. In this thesis an approach is introduced that reutilizes insights of the Semantic Web research to address this issue. As a result, the major contributions are the definition of a description language for federated infrastructures, the design of a reusable architecture for federated resource management, and the implementation of a proof of concept.
    Alexander Willner - Daniel Nehls (Substitute) (TU Berlin)