Seminar: Trends in Distributed Systems

Lecturer Prof. Dr. Christian Becker
Coordinator M.Sc. Jens Naber, M.Sc. Maximilian Roth
Type Seminar (SM 453 for Bachelor & IS 722 for Master students)
ECTS SM 453: 4/5 ECTS (depending on "Prüfungsordnung" [Examination Regulations])
IS 722: 6 ECTS for all. According to the current version of the "Prüfungsordnung" (Examination Regulations), M.Sc. Business Informatics gain 4 ECTS for the seminar and 2 ECTS for the key qualification "Scientific Research". If the old version of the "Prüfungsordnung" (Examination Regulations) applies for your studies, you will get an appropriate amount of ECTS.
ID-Number SM 453 (Bachelor seminar; former: SM 441)
IS 722 (Master seminar)
Prerequisites basic knowledge in information technology
Course Language English
Form of Assessment Conference style seminar (see details below)
Application process You may register via our Opens external link in new windowonline registration tool only (accessible inside the university network per VPN only).
We will not consider registrations by e-mail or incomplete data in the registration tool.
Topics will be assigned to students before the kick-off. Deregistrations are only accepted until September 17th, 2018, noon. Application Deadline: September 14th, 2018, noon.
Applicable for SM 453/441 (former): B.Sc. Business Informatics
IS 722: M.Sc. Management, M.Sc. Business Education, M.Sc. Business Informatics, Diploma Business Administration, Diploma Business Education
Acceptance notification September 14th, 2018
Kick-off session September 17th, 2018 at 13:45 pm - 15:15 am, L 15, 1-6 Room 714-715
Deadlines Seminar paper: November 2nd, 2018 at noon
Reviews: November 12th, 2018, noon
Camera ready paper: November 19th, 2018, noon
Final Presentation November 22nd and 23rd, 2018 at L15, 1-6 Room 714-715

Conference style seminar

This seminar is organized in a scientific conference style. All accepted participants must write a scientific paper about the assigned topics and submit those papers until the first draft deadline. After that, the paper review phase starts and each paper will be assigned to at least two other participants who have to review the papers of two or three other authors. After the review phase the reviews must be submitted to the supervisors, which distribute them to the paper authors. After that, the authors have time to improve their papers based on the feedback from the reviews, before they need to hand in their camera ready (final) version of the paper. At the end of the semester the "conference" (final presentations) will take place.

The grading is divided into different parts: The first part is the camera ready version of the seminar paper. This is the most important part and it is weighted with 50% of the overall grade. Second, the reviews for the other authors  are weighted with 20%. It is crucial to look at the work of others with a critical eye and to give constructive feedback. The last grading criterion is the presentation at the "conference" and the participation during the discussions (30%). 

If the assigned topic contains an implementation part the students don't need to participate in the review phase. For these topics the grading is divided into 70% for the implementation and 30% for the presentation and participation during the discussion.

Students have to pass each part separately. Attendance at the kick-off session and the final presentation session is mandatory.

For all papers the IEEE manuscript template must be used. Bellow we offer a customized version of the template (page numbering is already included).

Seminar topics

In this semester's seminar (HWS 2018), we offer the following topics.

 

Overview of Self-Adaptive Network Technologies

Supervisor: Martin Pfannemüller

Self-adaptive systems modify their behavior at run-time to maintain their performance after changes in the system resources or the environment. This is called adaptation. Therefore, an adaptation logic (AL) monitors the system resources and the environment, analyzes the monitored results, develops change plans, and executes them. This feedback loop can also be used to adapt networks at runtime in order to optimize, e.g., throughput or latency. The concepts of autonomic network management or policy-based network management are two possibilities to change the network behavior.

The objective of this literature-based seminar paper is to present and compare different self-adaptive network management systems. Additionally, the paper should refer to and embed Software-Defined Networking (SDN) as a current topic for network management into the picture of self-adaptive networks.

 

Distributed MAPE-K in Networked Cyber-Physical Systems

Supervisor: Martin Pfannemüller

Self-adaptive Systems modify their behavior at run-time to maintain their performance after changes in the system resources or the environment. This is called adaptation. Therefore, an adaptation logic (AL) monitors the system resources and the environment, analyzes the monitored results, develops change plans, and executes them. ALs typically follow the MAPE-K architecture. Multiple distribution patterns for this architecture exist. Cyber-physical systems (CPS), which are related to the Internet of Things, such as autonomous vehicles typically should be able to adapt themselves to new execution contexts. Thus, they need self-adaptive behavior. Additionally, they typically have to communicate with other system entities, so the networking capabilities are also very important.

The goal of this seminar work is to present an overview of CPS and IoT systems which use a distributed adaptation logic with network functionalities, e.g., following the MAPE-K architecture. The different adaptation logics should be discussed and categorized using a taxonomy.

 

Data Scheduling Strategies for Distributed Computing

Supervisor: Martin Breitbach

Especially for applications running on smartphones or smartwatches, computing power is still a scarce resource today. Distributed computing is one idea to overcome the devices’ limitations. Applications offload tasks to remote devices or the cloud via the network instead of running them locally.

Data scheduling handles the strategic distribution of input data such as images, videos, or text files in the distributed computing system to make offloading tasks as fast and reliable as possible. The goal of this seminar thesis is to provide an overview of data scheduling strategies for distributed computing based on a thorough literature review.

 

Meta-data Acquisition in Distributed Computing Systems

Supervisor: Martin Breitbach

Especially for applications running on smartphones or smartwatches, computing power is still a scarce resource today. Distributed computing is one idea to overcome the devices’ limitations. Applications offload tasks to remote devices or the cloud via the network instead of running them locally.

To effectively decide where to put tasks and data in a distributed computing system, it is important to collect meta-data. Examples for meta-data are the processing and storage capabilities of the available devices or the network topology. The objective of this seminar work is to identify strategies and protocols to perform meta-data acquisition in a distributed computing setting.   

 

Garbage Collection in Distributed Computing

Supervisor: Dominik Schäfer

Especially for applications running on smartphones or smartwatches, computing power is still a scarce resource today. Distributed computing is one idea to overcome the devices’ limitations. Applications offload tasks to remote devices or the cloud via the network instead of running them locally.

In many typical distributed computing use cases such as machine learning or image processing, devices exchange data of substantial size as input files or results. After a while, data may become outdated or unused. The aim of this seminar thesis is to identify garbage collection algorithms suitable for detecting and deleting unnecessary data in the system.

 

Identification of Access Barriers to Technology

Supervisor: Janick Edinger

Technological advances have made our lives more comfortable than ever before. Mobile devices, ubiquitous fast Internet access, and wearable sensor devices have changed the way we live, work, and communicate. These technologies do not only supplement existing areas of human life but also create new fields of entertainment, work, and social interaction. However, it is easy to forget that many people are excluded from these technologies for multiple reasons, including disabilities, diseases, or age-related limitations.

The objective of this seminar work is to identify different access barriers to technology and categorize them. It should be discussed why these barriers exist and how they can be tackled.

 

Unobtrusive Stress Measurement with Wearable Devices

Supervisor: Janick Edinger

In our everyday lives we are exposed to multiple stressful situations that might affect our health. To learn more about these situations and their effect on the human body, an accurate measurement of stress is crucial. For this purpose, it is necessary to first define what stress is and second, how it can be measured. Wearable devices allow for an unobtrusive way to measure physiological signs of stress in human subjects.

The objective of this seminar work is to identify physiological signs of stress and find solutions for a reliable and unobtrusive measurement of these factors. Therefore, a literature and market research should be performed.

 

A survey on privacy-preserving biometrics

Supervisor: Patricia Arias Cabarcos

New biometric technologies for authentication are being developed as alternative to passwords. The more promising are behavioral-based approaches that collect user patterns (e.g., keystrokes) for implicit and continuous identification. But this kind of mechanisms, as well as more traditional biometrics (e.g., fingerprint), require complex infrastructures or come with costly computations, so they are frequently outsourced. This raises privacy concerns, because biometric information is disclosed to a third party.

The goal of this seminar work is to research, categorize and analyze works dealing with privacy-preserving biometrics.

 

A survey on brainwave-based authentication

Supervisor: Patricia Arias Cabarcos

Electroencephalography (EEG) is the recording of electrical activity occurring in the brain, which has been repeatedly demonstrated to emit voltage fluctuations on a continuous basis. These fluctuations constitute a reflection of the on-going brain dynamics and vary from subject to subject. Grounded on this evidence, EEG signals (brainwaves) have been proposed as a way to identify users, creating an authentication mechanism alternative to passwords.

The goal of this seminar work is to survey, categorize and analyze EEG-based authentication proposals.

 

A meta-survey on IoT security challenges

Supervisor: Patricia Arias Cabarcos

The Internet of Things (IoT) creates new opportunities for direct integration of the physical world into computer-based systems, resulting in more efficient, intelligent, autonomous applications that can improve sectors such as commerce, industry, and education. One of the key issues for IoT technologies to succeed is security, reason for which the research community has extensively analyzed security and privacy challenges.

Given the high numbers of surveys on IoT security, the goal of this seminar work is to conduct a meta-survey (survey of surveys) to synthesize security challenges, update their importance and coverage by state-of-the-art countermeasures, and offer a consolidated view to serve as an entry point for researchers to approach the field in an effective way.

 

Trends in Pervasive Healthcare

Supervisor: Maximilian Roth

Pervasive systems consist of numerous connected and always available computing devices weaved into our everyday's life, such as smart health or smart home systems. They are able to automatically adapt to the context by changing system parameters or altering the context, e.g., turn lights on when somebody enters the room. In healthcare, such systems can be used for several application areas, such as health monitoring, data access, telemedicine, or health awareness.

The goal of this work is to elaborate and present trends of pervasive computing in the healthcare sector.

 

What defines good Visual Scripting?

Supervisor: Jens Naber

Today in many use cases, like home automation or workflow management, the developers try to empower the end user through Visual Scripting. With the help of a visual representation it is possible to give untrained users configuration options, which were formerly reserved for expert users. But to achieve this it is a crucial requirement to make the Visual Scripting Language intuitive and easy to use.

The goal of this work is to identify and discuss the theory behind good Visual Scripting Languages.

 

Survey on coordination principles in distributed systems

Supervisor: Jens Naber

In distributed systems it is often required to coordinate between the different peers. This could be to agree on a configuration, which peer is part of a specific group, or who is responsible for handling specific events. The task of coordination such a system is quite easy, if you have a central instance who you can make responsible for handling it, but it gets increasingly difficult with higher degrees of distribution.

The goal of this work is to identify and discuss the different coordination strategies for distributed systems.

 

Blockchain Tokenization of Energy

Supervisor: Benedikt Kirpes

Blockchain technology offers huge potential to disrupt the way energy is traded and paid for in the smart energy system of the future. Different approaches exist for “tokenizing” energy (= mapping energy units to a blockchain currency or tokens). Energy tokens can be used for different use cases, e.g. payment of electric vehicles charging transactions or direct peer-to-peer energy trading between neighbors.

The objective of this seminar work is to identify, compare and discuss relevant approaches and technologies for energy tokenization by conducting a literature review and market analysis.

 

Communication Technologies for Smart Energy Infrastructure

Supervisor: Benedikt Kirpes

Smart Grid and smart energy approaches need information and communication technologies which are capable of meeting the arising requirements for information exchange between devices. Infrastructure devices comprise Advanced Metering Infrastructure (AMI) like smart energy meters, intelligent charging stations or connected cars, mainly electric vehicles. Efficient and reliable communication between components is a necessity for a connected, coordinated and dynamic energy system, enabling promising approaches such as blockchain-based energy trading.

The objective of this seminar work is to identify and discuss relevant technologies, protocols and data formats by conducting a thorough literature review.

 

Mobile Gaze Estimation based on Computer Vision

Supervisor: Anton Wachner

We live in a mobile world in which manual input still is the predominant way to interact with a mobile device. Unobtrusive eye-based input methods are an emerging trend in human-computer interaction but most of the recommended systems still rely on dedicated hardware configurations such as devices equipped with an infra-red light module. Computer vision-based techniques exist that only need the raw image data as input and allow gaze estimation based on the pure camera input frames. However, accurate gaze estimation is still challenging in mobile scenarios as environmental factors vary such as lighting conditions and the position of the device relative to the user’s face.

The objective of this seminar work is to identify and categorize computer vision-based approaches for gaze estimation based on a thorough literature review. Further, implications for the applicability of selected approaches in mobile settings should be discussed.

 

Facial Feature Extraction based on Image Data

Supervisor: Anton Wachner

Today, manual input still is the predominant way to interact with a mobile device. Unobtrusive eye-based human-computer interfaces are an emerging trend in human-computer interaction and accurate gaze estimation is one of the key challenges. Frameworks exist that provide reliable face detection as well as facial landmarks detection based on raw image data. Image processing steps are performed to extract features that are used to localize faces within a frame. Once a face region is identified, a landmarks detection algorithm is applied to mark regions of interest such as the contours of the eyes, nose, and mouth. The information about the distribution of the landmarks around the eyes and the nose allows robust head pose estimation that in turn facilitates more precise gaze estimation. In real-world scenarios face detection fails if the captured portion of the face is too small. Therefore, methods are required that detect distinct regions of interest within a frame even if only relevant parts of the face are captured.

The objective of this seminar work is to find and review existing facial feature extraction approaches based on a thorough literature review. Further, the applicability in mobile context should be discussed.

 

***Implementation*** Head Pose Estimation on Mobile Devices

Supervisors: Anton Wachner

Today, manual input still is the predominant way to interact with a mobile device. Unobtrusive eye-based human-computer interfaces are an emerging trend in human-computer interaction. Reliable head pose estimation helps to predict where the user is looking. In the literature, there are many approaches for head pose estimation that can be adapted for mobile settings.

The goal of this seminar work is to identify head pose estimation algorithms and to select one of them for implementation in C/C++ or Java. The algorithm should be implemented and evaluated as part of this seminar thesis.

Note: This seminar topic includes implementation and its structure differs from the other topics. Please be aware that implementation in C/C++ will be an important part of this seminar work.