Seminar

Driven by the requirements of innovative applications and services, the past few years have produced new technologies for networked or distributed systems. In the Internet of Things, a large number of devices and everyday objects equipped with sensors and actuators are connected to the internet and communicate through mostly wireless communication technologies (e.g., BLE, ZigBee, 6LoWPAN, LoRaWAN). In one of its sub-domains, the industrial Internet of Things (IIoT, Industry 4.0), machines, tools, transport equipment, etc. are networked.
Virtualisation (e.g., NFV) and "software-defined" systems (e.g., SDN), increase the flexiblity and efficiency of distributed systems through dynamic adaptation and scaling.
Driven by the popularity of the Bitcoin system, distributed ledger technologies and related concepts such as smart contracts have been developed, which are not only the foundations of electronic currencies, but can also support any application in which a consensus between different parties must be reached and documented. 
Another focus has been the reduction of latency in networked and distributed systems, e.g. by using nearby edge and fog computing resources in addition to the remote cloud, or by using optimised communication protocols to., for instance, rapidly connect client and server. 
In addition to stationary networks, mobile (5G) communication technologies and systems have developed rapidly. For example, the Covid-19 tracing application uses mobile devices to track contacts. This method, known as crowdsensing, can be used more generally to collect large amounts of geographically distributed sensor data. 

This seminar will discuss a wide range of current technologies, protocols and standards that enable the above networked and distributed applications and services.

Organization: The seminar is organized in the style of a scientific conference. Following the submission of a written paper on the assigned topic, students write reviews for other seminar papers and participate in a final presentation session where they present their work and discuss the work of others. Attendance at the kick-off and final presentation session is mandatory.

Formal prerequisites: Successful completion of at least 1 course at the department of Distributed Systems.

TSN

Time-Sensitive Networks (TSN) refer to a set of networking standards developed by the IEEE to ensure reliable, low-latency, and deterministic data transmission across Ethernet-based networks. They are designed to handle real-time communication for critical applications in industrial automation, autonomous vehicles, and audio/video streaming. TSN achieves this by incorporating features such as time synchronization, traffic shaping, and reliability enhancements to prioritize and guarantee the delivery of time-critical data. It builds upon the IEEE 802.3 Ethernet technology, enabling it to meet the stringent safety and performance requirements of complex systems. With TSN, industries can converge both real-time and non-real-time data in a single network infrastructure, thus enhancing efficiency and reducing costs.

Topic 1: Traffic Shaping in TSN

Supervisor: Heiko Geppert

Traffic shaping within TSN plays a critical role in managing network resources by prioritizing time-sensitive data, reducing latency, and preventing congestion. The IEEE standards define several traffic shapers (Credit-Based Shaper, Time-Aware Shaper, and Asynchronous Traffic Shaper) to achieve different types of guarantees and reliability for real-time applications.

In the seminar, we will take a deep dive into these traffic shaping methods, highlighting their features and benefits. We will also explore their interaction with other TSN features, including per-stream filtering and policing.

Integration of 5G into TSN

The integration of 5G into TSN enables use cases with system components that require stringent communication guarantees and mobility. For example, an industrial control application may specify an end-to-end requirement with regard to latency, jitter, and reliability. To provide formal guarantees for such applications, 5G/TSN research focuses on the following areas:

Topic 2: Time Synchronization

Supervisor: Lucas Haug

Accurate clock synchronization is essential in networks requiring time-critical data transfer, where even minimal clock deviations can lead to significant operational errors. Given that the internal clocks of network devices inherently drift apart over time, clock synchronization is indispensable for reliable and error-free operations. In wired Ethernet networks, protocols like the Generalized Precision Time Protocol (gPTP), defined by the IEEE 802.1AS standard, are widely implemented. However, in emerging network domains – including converged 5G/TSN networks – time synchronization remains an active area of research. In the seminar, we will have an in-depths look at different aspects of time synchronization protocols in traditional Ethernet networks as well as in the novel converged 5G/TSN networks.

Topic 3: End-to-End Scheduling and 5G Resource Allocation

Supervisor: Lucas Haug, Simon Egger

Contrasting traffic shaping in wired TSN, scheduling in 5G-TSN must account for significant packet delays and packet delay variations which are induced by the 5G system. 5G packet delay variations are often in the range of milliseconds and orders of magnitude larger than the timing inaccuracies in wired TSN. As a result, scheduling in wireless TSN requires a careful analysis with respect to robustness and scalability. In the seminar, we will review state-of-the-art scheduling techniques from related domains (e.g., job-shop scheduling, queueing theory, multi-hop scheduling, and 5G resource allocation) and analyze their applicability to 5G/TSN.

Coordination and fault tolerance

Topic 4: Fault Tolerance and Impossibility Results in Distributed Systems

Supervisor: Simon Egger

To ensure the correct operation of higher-layer applications, distributed systems must specify their degree of fault tolerance (e.g., tolerance against crash faults, omission faults, or Byzantine faults). Within this research area, there exist a surprising number of impossibility results that bound the degree of fault tolerance that can be reached in different network settings. One such result is the prominent FLP theorem which shows that no agreement protocol can guarantee reaching a consensus in asynchronous networks with a single crash fault.

In the seminar, we will explore well-established techniques and research trends to achieve fault tolerance, with a special focus on real-time networks. Additionally, we will review impossibility results in this research area with the objective of analyzing the limitations that they impose on modern networking technologies.

Topic 5: Concurrency Control

Supervisor: Simon König

Concurrency and parallelism help programs utilize the available system resources optimally to speed up their calculations. In this seminar, we focus on controlling concurrent access to data and ensuring that the system remains in a consistent state. This comprises both distributed coordination and local scheduling. In the distributed setting, we investigate concurrency control algorithms and mechanisms for fault-tolerance. These ensure that distributed systems coordinate data access with low overhead to enable high throughput without sacrificing consistency. Locally, applications utilize memory optimizations that need to be carefully constructed to ensure consistency while maintaining scalability. Well-designed distributed systems combine all these optimizations and techniques.

In the seminar, we will investigate exciting new mechanisms, optimization techniques, and algorithms for concurrency control that build tomorrow’s large-scale applications.

Blockchain

Topic 6: Blockchain Privacy and Scalability

Supervisor: Lukas Epple

Privacy and scalability are essential components of modern blockchain systems. In the seminar, we will focus on advanced cryptographic techniques that enhance transaction anonymity and innovative solutions that improve blockchain efficiency without compromising security.

We will explore how privacy-focused cryptocurrencies employ methods like ring signatures, stealth addresses, and confidential transactions to keep transaction details anonymous and untraceable. Additionally, we will investigate how hash-based accumulator structures can compress the set of unspent transaction outputs (UTXOs), enabling more efficient blockchain verification by reducing the amount of data needed for transaction validation.

Well-designed blockchain systems combine these privacy and efficiency optimizations. In the seminar, we will explore the mechanisms, algorithms, and cryptographic techniques that underpin privacy and scalability in blockchain networks and learn how they contribute to secure, private, and efficient decentralized platforms.

Edge Computing

Topic 7: Edge Intelligence

Supervisor: Michael Schramm

With mobile communication technologies evolving constantly, edge computing theory and techniques have attracted an increasing interest from researchers and engineers around the world. Edge computing can help accelerate content delivery and reduce network load by communicating with nearby edge nodes instead of the cloud.

Another hot topic in computer science is new artificial intelligence (AI) applications, made possible by breakthroughs in deep learning and improvements in hardware architectures. However, the billions of data bytes that are generated at the network edge cannot all be transmitted to a central cloud server hosting the AI hardware infrastructure. This leads to a strong demand to bring AI to the edge and thus reduce the communication overhead, improve privacy, and reduce latency.

In the seminar, we will review scientific publications on Edge Intelligence. The goal is to describe different types of Edge Intelligence and compare them to cloud solutions. Both intelligence at the edge (i.e., making AI available on resource constrained devices) and intelligence for the edge (i.e., using machine learning techniques to optimize properties such as latency, Quality of Service, or energy consumption) are of interest. We will explore problems that scientific research currently addresses and discuss applications of Edge Intelligence.

Enabling technologies for IoT/Industry 4.0

Topic 8: WLAN

Supervisor: Jona Herrmann

Wireless Local Area Networks (WLAN), aka WiFi, are networks based on the IEEE 802.11 standard. These networks are essential for mobile devices, as they provide wireless access to the Internet – including in public places or on trains and buses. Additionally, they are an essential component of smart homes, where they facilitate easy communication between the many different connected devices. WLANs are also an integral part of industry, especially in Industry 4.0. An obvious advantage is that the communication is wireless, which makes it very flexible as there is no need for cabling.

In the seminar, we will take a deep dive into various aspects of WLANs.

Topic 9: Integration of Buildings into the Smart Grid

Supervisor: Sonja Klingert

The EU strategy of the “new Green Deal” targets a carbon neutral Europe until 2045. This requires the electricity grid to drastically increase its energy efficiency and amount of renewable energy resources, while considering its integration with other energy carriers. The challenges for the electricity grid are particularly high, as it physically needs to be balanced at all times. These challenges are multiplied by an increased share of small, variable, and less predictable decentralized energy generation alongside a continuous electrification of demand. This calls for increased flexibility in the electricity system.

A wide range of consuming devices and entities on the demand side can potentially serve as assets for flexibility – batteries and EVs being among the most cited. The flexibility contribution of building – viewed as a collection of consuming devices – can be optimized through concertation of usage, charging processes, electricity injection into the grid. Current flexibility schemes in buildings, however, have several shortcomings. First, the number of flexible assets they address is quite limited, due to the prevalence of legacy systems and low adoption of smart solutions (i.e., Building Energy Management Systems and IoT/smart devices). Second, although the set of core services that buildings provide to their occupants is rather well-defined and stationary, technologies and technical equipment required to implement them is heterogenous and varies over time. Thus, energy system operators struggle to maximally incorporate and benefit from demand side flexibility services.

Self-adaptation

Topic 10: Managing Uncertainty in Self-adaptive Systems

Supervisor: Michael Matthé

Modern software systems are becoming increasingly complex and beyond human control. One solution to this problem is self-adaptive systems. Self-adaptive systems monitor their environmental context and, in their simplest implementation, make configuration decisions based only on the monitored data. However, there are some uncertainties in this process: First, the monitored data may be incomplete or inaccurate. Second, it may be beneficial to reason about expected monitoring data. Making predictions about the future behavior of the environment enables proactive adaptation. However, making good adaptation decisions requires an accurate way of predicting the future context. In addition, the execution of the adaptation may incur costs (e.g., latency, energy consumption), which can introduce uncertainty about the actual benefit of the adaptation – depending on the current system context.

In the seminar, we will look at specific research areas that explore approaches to handle one of the aforementioned dimensions of uncertainty in self-adaptive systems.

Topic 11: Testing and Benchmarking Self-adaptive Systems

Supervisor: Melanie Heck

In order to evaluate novel techniques for self-adaptive systems, the software is often tested on web applications running on real web servers. This makes the evaluation highly challenging, as the deployment of the systems is neither easy nor cheap. Additionally, an exact replication of runtime conditions to compare different adaptation approaches is basically impossible due uncontrollable contextual variables. In order to obtain meaningful results, a large amount of test data must be collected, making the process extremely time consuming. To address these issues, the self-adaptive systems community has promoted the development of simulation software, which is now increasingly being used for evaluation. This allows researchers to compare their approaches to state-of-the art solutions without also having to re-run the benchmark evaluation. The simulators vary widely in multiple aspects, including their complexity, ease of configuration, and the parameters that can be manipulated in order to test specific aspects of self-adaptation.

In the seminar, we will examine the landscape of both physical test environments and simulators with the aim of identifying their advantages and disadvantages for evaluating different aspects of self-adaptation.

Melanie Heck

The Internet of Everything (IoE), where virtually everything can now communicate through the Internet, and the increasingly demanding performance requirements of new technologies (e.g., cryptocurrencies) have driven the emergence of new computing paradigms for distributed systems. Scalability is now offered not only by centralized cloud providers, but also by edge computing systems, where geographically distributed servers provide computational resources at the edge of the network and, therefore, close to the end devices. This can significantly reduce latency for time-critical applications like vehicular networks. The advances in edge computing have led to the emergence of edge AI, where powerful AI algorithms are deployed at the edge, without relying on a remote cloud.

But distributed systems come with many challenges which requires a profound understanding of core principles in distributed computing. As pointed out by former Google Senior Vice President Urs Hölzl: “At scale, everything breaks ... Keeping things simple and yet scalable is actually the biggest challenge. It's really, really hard.“ This is especially true for dynamic and uncertain environments that we are facing, for instance, in smart buildings or smart energy systems. Self-adaptation is one of the key mechanisms for coping with increasingly large and dynamic systems, often by using machine learning techniques (GNN, reinforcement learning). Challenges that come with distributed storage systems include consistency and scalability.

Another hot topic, especially in the context of 5G and the development of future 6G networks, is Time Sensitive Networking (TSN), which defines a set of standards to enable reliable, deterministic real-time communication in Ethernet networks. These standards target, among others, time synchronization and traffic shaping/scheduling approaches for both event-based and time-triggered traffic.

In this seminar, we take a deep dive into specific concepts of distributed and context-aware systems that tackle the above challenges.  The topics will be published on the department’s website and are assigned according to a standardized procedure as explained during the kick-off.

Organization: The seminar is organized in the style of a scientific conference. Following the submission of a written paper on the assigned topic, students write reviews for other seminar papers and participate in a final presentation session where they present their work and discuss the work of others. Attendance at the kick-off and final presentation session is mandatory.

Formal prerequisite: Successful completion of at least 1 Master-level course at the department of Distributed Systems.

TSN

Time-Sensitive Networks (TSN) refer to a set of networking standards developed by the IEEE to ensure reliable, low-latency, and deterministic data transmission across Ethernet-based networks. They are designed to handle real-time communication for critical applications in industrial automation, autonomous vehicles, and audio/video streaming. TSN achieves this by incorporating features such as time synchronization, traffic shaping, and reliability enhancements to prioritize and guarantee the delivery of time-critical data. It builds upon the IEEE 802.3 Ethernet technology, enabling it to meet the stringent safety and performance requirements of complex systems. With TSN, industries can converge both real-time and non-real-time data in a single network infrastructure, thus enhancing efficiency and reducing costs.

Topic 1: Traffic Shaping in TSN

Supervisor: Heiko Geppert

Traffic shaping within TSN plays a critical role in managing network resources by prioritizing time-sensitive data, reducing latency, and preventing congestion. The IEEE standards define several traffic shapers (Credit-Based Shaper, Time-Aware Shaper, and Asynchronous Traffic Shaper) to achieve different types of guarantees and reliability for real-time applications.

In the seminar, we will take a deep dive into these traffic shaping methods, highlighting their features and benefits. We will also explore their interaction with other TSN features, including per-stream filtering and policing.

Integration of 5G into TSN

The integration of 5G into TSN enables use cases with system components that require stringent communication guarantees and mobility. For example, an industrial control application may specify an end-to-end requirement with regard to latency, jitter, and reliability. To provide formal guarantees for such applications, 5G/TSN research focuses on the following areas:

Topic 2: Time Synchronization

Supervisor: Lucas Haug

Accurate clock synchronization is essential in networks requiring time-critical data transfer, where even minimal clock deviations can lead to significant operational errors. Given that the internal clocks of network devices inherently drift apart over time, clock synchronization is indispensable for reliable and error-free operations. In wired Ethernet networks, protocols like the Generalized Precision Time Protocol (gPTP), defined by the IEEE 802.1AS standard, are widely implemented. However, in emerging network domains – including converged 5G/TSN networks – time synchronization remains an active area of research. In the seminar, we will have an in-depths look at different aspects of time synchronization protocols in traditional Ethernet networks as well as in the novel converged 5G/TSN networks.

Topic 3: End-to-End Scheduling and 5G Resource Allocation

Supervisor: Lucas Haug, Simon Egger

Contrasting traffic shaping in wired TSN, scheduling in 5G-TSN must account for significant packet delays and packet delay variations which are induced by the 5G system. 5G packet delay variations are often in the range of milliseconds and orders of magnitude larger than the timing inaccuracies in wired TSN. As a result, scheduling in wireless TSN requires a careful analysis with respect to robustness and scalability. In the seminar, we will review state-of-the-art scheduling techniques from related domains (e.g., job-shop scheduling, queueing theory, multi-hop scheduling, and 5G resource allocation) and analyze their applicability to 5G/TSN.

Coordination and fault tolerance

Topic 4: Fault Tolerance and Impossibility Results in Distributed Systems

Supervisor: Simon Egger

To ensure the correct operation of higher-layer applications, distributed systems must specify their degree of fault tolerance (e.g., tolerance against crash faults, omission faults, or Byzantine faults). Within this research area, there exist a surprising number of impossibility results that bound the degree of fault tolerance that can be reached in different network settings. One such result is the prominent FLP theorem which shows that no agreement protocol can guarantee reaching a consensus in asynchronous networks with a single crash fault.

In the seminar, we will explore well-established techniques and research trends to achieve fault tolerance, with a special focus on real-time networks. Additionally, we will review impossibility results in this research area with the objective of analyzing the limitations that they impose on modern networking technologies.

Topic 5: Concurrency Control

Supervisor: Simon König

Concurrency and parallelism help programs utilize the available system resources optimally to speed up their calculations. In this seminar, we focus on controlling concurrent access to data and ensuring that the system remains in a consistent state. This comprises both distributed coordination and local scheduling. In the distributed setting, we investigate concurrency control algorithms and mechanisms for fault-tolerance. These ensure that distributed systems coordinate data access with low overhead to enable high throughput without sacrificing consistency. Locally, applications utilize memory optimizations that need to be carefully constructed to ensure consistency while maintaining scalability. Well-designed distributed systems combine all these optimizations and techniques.

In the seminar, we will investigate exciting new mechanisms, optimization techniques, and algorithms for concurrency control that build tomorrow’s large-scale applications.

Blockchain

Topic 6: Blockchain Privacy and Scalability

Supervisor: Lukas Epple

Privacy and scalability are essential components of modern blockchain systems. In the seminar, we will focus on advanced cryptographic techniques that enhance transaction anonymity and innovative solutions that improve blockchain efficiency without compromising security.

We will explore how privacy-focused cryptocurrencies employ methods like ring signatures, stealth addresses, and confidential transactions to keep transaction details anonymous and untraceable. Additionally, we will investigate how hash-based accumulator structures can compress the set of unspent transaction outputs (UTXOs), enabling more efficient blockchain verification by reducing the amount of data needed for transaction validation.

Well-designed blockchain systems combine these privacy and efficiency optimizations. In the seminar, we will explore the mechanisms, algorithms, and cryptographic techniques that underpin privacy and scalability in blockchain networks and learn how they contribute to secure, private, and efficient decentralized platforms.

Edge Computing

Topic 7: Edge Intelligence

Supervisor: Michael Schramm

With mobile communication technologies evolving constantly, edge computing theory and techniques have attracted an increasing interest from researchers and engineers around the world. Edge computing can help accelerate content delivery and reduce network load by communicating with nearby edge nodes instead of the cloud.

Another hot topic in computer science is new artificial intelligence (AI) applications, made possible by breakthroughs in deep learning and improvements in hardware architectures. However, the billions of data bytes that are generated at the network edge cannot all be transmitted to a central cloud server hosting the AI hardware infrastructure. This leads to a strong demand to bring AI to the edge and thus reduce the communication overhead, improve privacy, and reduce latency.

In the seminar, we will review scientific publications on Edge Intelligence. The goal is to describe different types of Edge Intelligence and compare them to cloud solutions. Both intelligence at the edge (i.e., making AI available on resource constrained devices) and intelligence for the edge (i.e., using machine learning techniques to optimize properties such as latency, Quality of Service, or energy consumption) are of interest. We will explore problems that scientific research currently addresses and discuss applications of Edge Intelligence.

Enabling technologies for IoT/Industry 4.0

Topic 8: WLAN

Supervisor: Jona Herrmann

Wireless Local Area Networks (WLAN), aka WiFi, are networks based on the IEEE 802.11 standard. These networks are essential for mobile devices, as they provide wireless access to the Internet – including in public places or on trains and buses. Additionally, they are an essential component of smart homes, where they facilitate easy communication between the many different connected devices. WLANs are also an integral part of industry, especially in Industry 4.0. An obvious advantage is that the communication is wireless, which makes it very flexible as there is no need for cabling.

In the seminar, we will take a deep dive into various aspects of WLANs.

Topic 9: Integration of Buildings into the Smart Grid

Supervisor: Sonja Klingert

The EU strategy of the “new Green Deal” targets a carbon neutral Europe until 2045. This requires the electricity grid to drastically increase its energy efficiency and amount of renewable energy resources, while considering its integration with other energy carriers. The challenges for the electricity grid are particularly high, as it physically needs to be balanced at all times. These challenges are multiplied by an increased share of small, variable, and less predictable decentralized energy generation alongside a continuous electrification of demand. This calls for increased flexibility in the electricity system.

A wide range of consuming devices and entities on the demand side can potentially serve as assets for flexibility – batteries and EVs being among the most cited. The flexibility contribution of building – viewed as a collection of consuming devices – can be optimized through concertation of usage, charging processes, electricity injection into the grid. Current flexibility schemes in buildings, however, have several shortcomings. First, the number of flexible assets they address is quite limited, due to the prevalence of legacy systems and low adoption of smart solutions (i.e., Building Energy Management Systems and IoT/smart devices). Second, although the set of core services that buildings provide to their occupants is rather well-defined and stationary, technologies and technical equipment required to implement them is heterogenous and varies over time. Thus, energy system operators struggle to maximally incorporate and benefit from demand side flexibility services.

Self-adaptation

Topic 10: Managing Uncertainty in Self-adaptive Systems

Supervisor: Michael Matthé

Modern software systems are becoming increasingly complex and beyond human control. One solution to this problem is self-adaptive systems. Self-adaptive systems monitor their environmental context and, in their simplest implementation, make configuration decisions based only on the monitored data. However, there are some uncertainties in this process: First, the monitored data may be incomplete or inaccurate. Second, it may be beneficial to reason about expected monitoring data. Making predictions about the future behavior of the environment enables proactive adaptation. However, making good adaptation decisions requires an accurate way of predicting the future context. In addition, the execution of the adaptation may incur costs (e.g., latency, energy consumption), which can introduce uncertainty about the actual benefit of the adaptation – depending on the current system context.

In the seminar, we will look at specific research areas that explore approaches to handle one of the aforementioned dimensions of uncertainty in self-adaptive systems.

Topic 11: Testing and Benchmarking Self-adaptive Systems

Supervisor: Melanie Heck

In order to evaluate novel techniques for self-adaptive systems, the software is often tested on web applications running on real web servers. This makes the evaluation highly challenging, as the deployment of the systems is neither easy nor cheap. Additionally, an exact replication of runtime conditions to compare different adaptation approaches is basically impossible due uncontrollable contextual variables. In order to obtain meaningful results, a large amount of test data must be collected, making the process extremely time consuming. To address these issues, the self-adaptive systems community has promoted the development of simulation software, which is now increasingly being used for evaluation. This allows researchers to compare their approaches to state-of-the art solutions without also having to re-run the benchmark evaluation. The simulators vary widely in multiple aspects, including their complexity, ease of configuration, and the parameters that can be manipulated in order to test specific aspects of self-adaptation.

In the seminar, we will examine the landscape of both physical test environments and simulators with the aim of identifying their advantages and disadvantages for evaluating different aspects of self-adaptation.

Melanie Heck