Emerging applications such as autonomous driving, virtual reality, and smart factories place greater demands on the Quality of Service of existing network infrastructure, particularly radio networks. The current 5th and new 6th generation of cellular networks aim to meet these requirements and provide ubiquitous connectivity to devices with diverse demands. These networks comprise a control plane and a user plane. While the control plane is responsible for managing the network and its devices, the user plane forwards data and directly influences the experienced Quality of Service. A key network function in the user plane is the User Plane Function (UPF), which forwards packets between cellular network devices and the data network, such as the Internet or an edge data center. However, the extent to which existing UPF implementations can provide sufficient Quality of Service for emerging applications remains largely unexplored. In this work, we analyze and compare various UPF implementations from both theoretical and practical perspectives. We consider both software-based and hardware-accelerated implementations and compare them in terms of performance and latency under load. The setup enables up to 10,000 subscriber sessions while enforcing QoS mechanisms such as rate limiting. The evaluation demonstrates that three of the four investigated UPFs provide QoS enforcement, while their latency behavior differs by orders of magnitude depending on the employed technology.

Quantum networks are becoming increasingly important because of advancements in quantum computing and quantum sensing, such as recent developments in distributed quantum computing and federated quantum machine learning. Routing entanglement in quantum networks poses several fundamental as well as technical challenges, including the high dynamicity of quantum network links and the probabilistic nature of quantum operations. Consequently, designing hand-crafted heuristics is difficult and often leads to suboptimal performance, especially if global network topology information is unavailable. In this paper, we propose RELiQ, a reinforcement learningbased approach to entanglement routing that only relies on local information and iterative message exchange. Utilizing a graph neural network, RELiQ learns graph representations and avoids overfitting to specific network topologies – a prevalent issue for learning-based approaches. Our approach, trained on random graphs, consistently outperforms existing local information heuristics and learning-based approaches when applied to random and real-world topologies. When compared to global information heuristics, our method achieves similar or superior performance because of its rapid response to topology changes.

Large-scale packet queueing and scheduling is the basis for today's Quality of Service (QoS) in computer access networks, especially to achieve guaranteed high throughput and low latency. The throughput of a single network function, implementing the QoS functionality for a large number of customers, is in the range of hundreds of gigabits or even more. Therefore, a good performance of the underlying hardware is mandatory. While highly-performant fixed-function ASICs offer sufficient functionality for most data center use cases, as of today, they cannot support all functionality required for access networks, including QoS-aware packet queuing. In this paper, we first analyze mobile and residential Internet access requirements from an Internet service provider perspective, focusing on the QoS-aware packet queueing needs. Considering this analysis, we present a universal and generic FPGA design for high-performance packet queueing and scheduling. Our evaluation results show that FPGAs can be used to implement a deterministic QoS packet queueing system with high performance. This concept can extend today's programmable networking ASICs with the desired functionality as an offloaded “sidecar”.

There is growing interest in point cloud content due to its central role in the creation and provision of interactive and immersive user experiences for extended reality applications. However, it is impractical to stream uncompressed point cloud sequences over communication networks to end systems because of their high throughput and low latency requirements. Several novel compression methods have been developed for efficient storage and adaptive delivery of point cloud content. However, these methods primarily focus on data metrics and neglect the influence on the actual Quality of Experience (QoE). In this paper, we conduct a user study with 102 participants to analyze the QoE of point cloud sequences and develop a QoE model that can enhance the quality of point cloud content distribution under dynamic network conditions. Our analysis is based on user opinions regarding two representative point cloud sequences, three different frame rates, three viewing distances, and two state-of-the-art point cloud compression libraries, Draco and V-PCC. The results indicate that the proposed models can accurately predict the users' quality perception, with frame rate being the most dominant QoE factor.

Handing in a paper or exercise and merely receiving "bad" or "incorrect" as feedback is not very helpful when the goal is to improve.Unfortunately, this is currently the kind of feedback given by many Automatic Short Answer Grading (ASAG) systems.One of the reasons for this is a lack of content-focused elaborated feedback datasets.To encourage research on explainable and understandable feedback systems, we present the Short Answer Feedback dataset (SAF).Similar to other ASAG datasets, SAF contains learner responses and reference answers to German and English questions.However, instead of only assigning a label or score to the learners' answers, SAF also contains elaborated feedback explaining the given score.Thus, SAF enables supervised training of models that grade answers and explain where and why mistakes were made.This paper discusses the need for enhanced feedback models in real-world pedagogical scenarios, describes the dataset annotation process, gives a comprehensive analysis of SAF, and provides T5-based baselines for future comparison. 1

Large-scale packet queueing and scheduling is the basis for today’s Quality of Service (QoS) in computer access networks, especially to achieve guaranteed high throughput and low latency. While highly performant fixed-function ASICs offer sufficient functionality for most data center use-cases, as of today, they cannot support all functionality required for access networks, e.g., QoS-aware packet queueing.In this poster, we present an FPGA-based architecture for network packet queueing optimized for residential and mobile Internet access networks.

Hardware acceleration of network functions is essential to meet the challenging Quality of Service requirements in nowadays computer networks. Graphical Processing Units (GPU) are a widely deployed technology that can also be used for computing tasks, including acceleration of network functions. In this work, we demonstrate how commodity GPUs, which do not provide any network interfaces, can be used to accelerate network functions. Our approach leverages PCIe peer-to-peer capabilities and allows the GPU to control the network interface card directly, without any assistance from the operating system or control application. The presented evaluation results demonstrate the feasibility of our approach and its performance of up to 10 Gbit/s, even for small packets.

Congestion control mechanisms in computer networks rely mainly on a feedback loop having a reaction time equal to the flow RTT. Reducing this feedback time helps the sender to react faster to changing network conditions such as congestion. In this work, we propose reverse-path congestion notification on top of programmable networking switches. Our approach can significantly lower the reaction time, such that the congestion control implementation can adapt much faster to changing network conditions. The proposed approach aims to work with current TCP implementations with no required changes to the communication endpoints. Last, we show how the presented approach could be realized by utilizing off-the-shelf programmable switches.

Cooperative Vehicular Maneuvering (CVM) significantly improves traffic safety and efficiency, but relies on wireless information exchange between cooperating vehicles. As wireless links are unreliable and the communication resources are limited, the performance of CVM degrades depending on the communication quality. We propose a novel model to represent the imperfectness of communication, formalize CVM at an intersection, and evaluate the performance degradation of CVM under varying communication conditions. From our analytical perspective, we can derive the requirements for maneuver coordination. Our results provide a valuable resource for the design of future communication technologies and dissemination strategies.

Advanced Driver Assistance Systems require a tremendous amount of sensor information to support the driver's comfort and safety. In particular, systems that provide (good) route options to a vehicle rely on information, such as traffic jams and road blockages, which is sensed by other (possibly distant) vehicles and distributed by a central server. This information is clearly dynamic and may be invalid by the time the vehicle arrives at the affected location. In this work, we develop an innovative approach to determine optimal routes (minimizing the costs like travel-time to their destination) for vehicles whose original route is adversely impacted by a (severe) road event. A set of recursive equations is developed that yields the optimal decision for each vehicle at each decision-point. Simulations show that our approach adapts to the considered event and finds routes of similar quality as a full-knowledge approach with limited communication overhead.

Performance evaluations for heterogeneous communication technologies in the area of V2X safety applications for either improvement, comparison or combination purposes are in general focusing on the realistic representation of the upper communication stack layers, but therefore - often for the sake of simplicity - reducing the radio propagation channel to a maximum range model. The impact and hence the importance to model the environment dependent propagation effects in a representative manner has already been stressed in the literature several times - but separately for ad-hoc or cellular systems and not under the consideration of V2X safety-beaconing applications. By combining a realistic heterogeneous radio propagation channel model with a state-of-the-art V2X communication stack, a representative performance comparison of safety-relevant beaconing applications for 802.11p single-hop broadcast (SHB) and LTE Geocast can be conducted. Our simulation results show that the effects caused by the radio propagation channel cannot be neglected as they significantly impact key communication performance metrics such as channel gain, packet error ratio (PER) and channel load, where we primarily focus on the latter one to give further research directions for an efficient dissemination of safety-relevant V2X beacons.

Direct ad hoc connections among mobile devices can be utilized for emergency communication in times when communication infrastructure is unavailable due to blackouts or natural disasters. However, the quality of the resulting mobile ad hoc network strongly depends on the number of devices involved. Thus, to sustain a fully functional emergency communication network, devices need to recharge using external resources (e.g., battery packs, solar panels). Access to these resources needs to be coordinated such that the overall network lifetime is increased. We propose an auction-based resource reservation protocol for a decentralized resource allocation service. Through an extensive simulation study we show that our approach (i) efficiently coordinates the competition for resources, leading to a network lifetime comparable to a global allocation approach (98.8%) and (ii) delays the time until the first node runs out of energy by a factor of 1.4 compared to related work. This increases the value of the emergency communication network significantly, aiding more people over extended periods of time.

As a step towards fully autonomous driving, Advanced Driver Assistance Systems provide convenience-and safety-related functions to drivers. In addition to data gathered by local sensors, these systems rely on events generated by other vehicles that need to be disseminated to a potentially large audience. Today, this geocast-functionality relies either on subscriptions covering certain areas (e.g., cities) or on individual route-based subscriptions. While the former exhibits suboptimal precision in filtering, the latter introduces significant complexity and assume that routes are known in advance. We propose a prediction-based assessment of the relevance of events without requiring prior route knowledge. Relevance is modeled based on the street network and spatio-temporal characteristics of events. We evaluate our approach in a realistic city setting, relying on the SUMO vehicular mobility simulator. Our first results show that relevance-aware information dissemination reduces the communication overhead by 68%, while at the same time achieving near perfect recall compared to route-based subscrintions.