SKSabrina Klos

Sabrina Klos

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Sabrina Klos has a total of 13 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Multi-Stakeholder Service Placement via Iterative Bargaining With Incomplete Information

    by Artur Sterz, Patrick Felka, Bernd Simon, Sabrina Klos, Anja Klein, Oliver Hinz, Bernd Freisleben

    Mobile edge computing based on cloudlets is an emerging paradigm to improve service quality by bringing computation and storage facilities closer to end users and reducing operating cost for infrastructure providers (IPs) and service providers (SPs). To maximize their individual benefits, IP and SP have to reach an agreement about placing and executing services on particular cloudlets. We show that a Nash Bargaining Solution (NBS) yields the optimal solution with respect to social cost and fairness if IP and SP have complete information about the parameters of their mutual cost functions. However, IP and SP might not be willing or able to share all information due to business secrets or technical limitations. Therefore, we present a novel iterative bargaining approach without complete mutual information to achieve substantial cost reductions for both IP and SP. Furthermore, we investigate how different degrees of information sharing impact social cost and fairness of the different approaches. Our evaluation based on the mobile augmented reality game Ingress shows that our approach achieves up to about 82% of the cost reduction that the NBS achieves and a cost reduction of up to 147% compared to traditional Take-it-or-Leave-it approaches, despite incomplete information.


  • An Online Context-Aware Machine Learning Algorithm for 5G mmWave Vehicular Communications

    by Gek Hong Sim, Sabrina Klos, Arash Asadi, Anja Klein, Matthias Hollick

    Millimeter-Wave (mmWave) bands have become the de-facto candidate for 5G vehicle-to-everything (V2X) since future vehicular systems demand Gbps links to acquire the necessary sensory information for (semi)-autonomous driving. Nevertheless, the directionality of mmWave communications and its susceptibility to blockage raise severe questions on the feasibility of mmWave vehicular communications. The dynamic nature of 5G vehicular scenarios and the complexity of directional mmWave communication calls for higher context-awareness and adaptability. To this aim, we propose an online learning algorithm addressing the problem of beam selection with environment-awareness in mmWave vehicular systems. In particular, we model this problem as a contextual multi-armed bandit problem. Next, we propose a lightweight context-aware online learning algorithm, namely fast machine learning (FML), with proven performance bound and guaranteed convergence. FML exploits coarse user location information and aggregates the received data to learn from and adapt to its environment. Furthermore, we demonstrate the feasibility of a real-world implementation of FML by proposing a standard-compliant protocol based on the existing architecture of cellular networks and the forthcoming features of 5G. We also perform an extensive evaluation using realistic traffic patterns derived from Google Maps. Our evaluation shows that FML enables mmWave base stations to achieve near-optimal performance on average within 33 mins of deployment by learning from the available context. Moreover, FML remains within ~5% of the optimal performance by swift adaptation to system changes (i.e., blockage, traffic).


  • Context-Aware Hierarchical Online Learning for Performance Maximization in Mobile Crowdsourcing

    by Sabrina Klos Nee Muller, Cem Tekin, Mihaela van der Schaar, Anja Klein

    In mobile crowdsourcing (MCS), mobile users accomplish outsourced human intelligence tasks. MCS requires an appropriate task assignment strategy, since different workers may have different performance in terms of acceptance rate and quality. Task assignment is challenging, since a worker's performance 1) may fluctuate, depending on both the worker's current personal context and the task context and 2) is not known a priori, but has to be learned over time. Moreover, learning context-specific worker performance requires access to context information, which may not be available at a central entity due to communication overhead or privacy concerns. In addition, evaluating worker performance might require costly quality assessments. In this paper, we propose a context-aware hierarchical online learning algorithm addressing the problem of performance maximization in MCS. In our algorithm, a local controller (LC) in the mobile device of a worker regularly observes the worker's context, her/his decisions to accept or decline tasks and the quality in completing tasks. Based on these observations, the LC regularly estimates the worker's context-specific performance. The mobile crowdsourcing platform (MCSP) then selects workers based on performance estimates received from the LCs. This hierarchical approach enables the LCs to learn context-specific worker performance and it enables the MCSP to select suitable workers. In addition, our algorithm preserves worker context locally, and it keeps the number of required quality assessments low. We prove that our algorithm converges to the optimal task assignment strategy. Moreover, the algorithm outperforms simpler task assignment strategies in experiments based on synthetic and real data.


  • Top Cited Papers
    Open Access

    Context-Aware Proactive Content Caching With Service Differentiation in Wireless Networks

    by Sabrina Muller, Onur Atan, Mihaela van der Schaar, Anja Klein

    Content caching in small base stations or wireless infostations is considered to be a suitable approach to improve the efficiency in wireless content delivery. Placing the optimal content into local caches is crucial due to storage limitations, but it requires knowledge about the content popularity distribution, which is often not available in advance. Moreover, local content popularity is subject to fluctuations, since mobile users with different interests connect to the caching entity over time. Which content a user prefers may depend on the user's context. In this paper, we propose a novel algorithm for context-aware proactive caching. The algorithm learns context-specific content popularity online by regularly observing context information of connected users, updating the cache content and observing cache hits subsequently. We derive a sublinear regret bound, which characterizes the learning speed and proves that our algorithm converges to the optimal cache content placement strategy in terms of maximizing the number of cache hits. Furthermore, our algorithm supports service differentiation by allowing operators of caching entities to prioritize customer groups. Our numerical results confirm that our algorithm outperforms state-of-the-art algorithms in a real world data set, with an increase in the number of cache hits of at least 14%.