Zum Inhalt springen
  • de
  • en
Menü
Technische Universität München
Technische Universität München

Advanced Seminar Embedded Systems and Internet of Things

Our kick-off meeting on the 22nd of April at 9:45 will be conducted in an online meeting, more information will follow soon with a mail to all accepted participants.

This semester we offer the following thirteen topics for the advanced seminar "Embedded Systems and Internet of Things":

  1. Run-Time Timing Analysis for Predictable Task Execution in Composable Distributed Systems
  2. Predictable Task Migration in Distributed Systems
  3. Transport Protocols, what is the Destination?
  4. Find Your Match, in the Network, but it is a Thing
  5. The Role of Butterfly Key Expansion with Regard to Privacy Enhanced Credential Management in the IoV
  6. Graph-based Authentication Schemes to Support Blockchain-Based Identity Management Systems
  7. WiFi Authentication and Key Generation for IoT
  8. Lattice-based Post-Quantum Signatures
  9. Performance Analysis Problems in TSN Networks
  10. Reconfiguration Problems in TSN Networks
  11. Runtime Monitoring / Verification for Cyber-Physical Systems
  12. Formal Specifications for Cyber-Physical Systems
  13. Using Artificial Intelligence Mechanisms to Implement Intrusion Response System

You will find the description of the topics below. Furthermore, we put reference papers for each topic as a starting point for your research. Your task for each topic will be to read and analyze several references, get an overview of the current state-of-the-art and summarize your findings in a paper-style report. Afterwards you will present your findings in a short talk in front of your fellow students.

During the seminar you will also learn how to conduct the research, how to write a scientific paper and how to present your work.

 

Topic Assignment

Due to the high interest in our seminar topics we use an application process to assign the topics.

If you are interested in one of the topics, please apply for a topic by writing an email to the supervisor of the topic. Express your interest and why you want to have that specific topic and why you think that you are most suitable for the topic.

Additionally, you should state a second topic that you would like to take, such that we can still find a topic for you if your primary choice is not available.

Deadline: We encourage you to apply until the 08.04.2020. We will assign topics on the 08.04.2020 based on the requests until then and mark which topics are taken. After this date, we will answer to requests within 3 days, assuming that there is enough motivation for the given topic. Once you are given the topic, we will ask for your confirmation.

Note: Applying for a topic in advance shows your interest but we can still choose another applicant. Generally we have 3-4 applicants per topic. Please think carefully if you are able to do the work required, we will have to say no to other students for you to take this topic. Generally, email clients remember the people you have communicated with.

 

Kick-off meeting

The kick-off meeting will be on the 22nd of April at 9:45. We will inform all accepted applicants in time how the meeting will be conducted. We ask all selected participants who have been assigned a topic to be present in the kick-off meeting. Please notify us in case you can not make it to the meeting, otherwise we will give your place to another applicant.

 

Topics

1. Run-Time Timing Analysis for Predictable Task Execution in Composable Distributed Systems

Time-critical automotive applications have to meet real-time deadlines. However, interference caused by tasks that are being executed on a shared resource make timing predictions difficult. Timing composability allows to independently analyze the execution time of tasks in a multitasking system. Composability can be improved by eliminating or minimizing resource interference. An approach to minimize interference is using temporal and spatial isolation schemes. An additional challenge is that in future automotive system the executed applications might change at run-time.  This requires to perform timing analysis at run-time to adapt the system accordingly. Your task is to investigate approaches that perform online timing analysis and can give timing guarantees at run-time using isolation schemes.

References:

Supervisor: Philipp Weiß

Topic unassigned

 

2. Predictable Task Migration in Distributed Systems

Dynamic resource managment techniques rely on task migration to adapt task mappings at run-time. However, migrating tasks at run-time imposes an additional timing overhead, which might interfere with real-time requirements. Your task is to analyze approaches which allow task migration and which consider real-time constraints during the migration process.

References:

Supervisor: Philipp Weiß

Topic assigned

 

3. Transport Protocols, what is the Destination?

The backbone of the Web is about receive a big update where the TCP-based HTTP/2 is being replaced by UDP-based QUIC protocol that is also being integrated into HTTP/3. It promises better performance and security, with the former being the goal of this seminar topic. The student shall investigate its performance benefit claims, while also analyzing its fundamental differences to the TCP-based HTTP/2 and HTTP/1.1. Additional comparisons to other protocols such as CoAP are also encouraged.

References:

Supervisor: Ege Korkan

Topic assigned

 

4. Find Your Match, in the Network, but it is a Thing

Service Discovery protocols are used in the Internet or local networks to find other devices such as servers who offer a service. The task is to analyze the state-of-the-art service protocols such as the zeroconf stack and compare them qualitatively and if possible quantitively. Additional comparisons to the CoRE Resource Directory are welcomed, as well as consideration of resource constrained IoT nodes.

References:

Supervisor: Ege Korkan

Topic unassigned

 

5. The Role of Butterfly Key Expansion with Regard to Privacy Enhanced Credential Management in the IoV

The holistic and increasingly complex infrastructure of the Internet of vehicles (IoV) absorbs and monitors the environment to provide assistance in driving decisions, navigation management, and traffic control. However, some of the processes lead to privacy and security violations. As a result, data repositories, sensors as well as human stakeholders require secure, efficient, and privacy preserving credential management procedures and systems which keep up with the increasing amount of scale of devices. For this topic and in the context of the IoV, you should compare how the unified butterfly key (UBK) expansion process enhances privacy of credential managment procedures compared to credential management procedures of traditional PKI systems.

References:

Supervisor: Jan Lauinger

Topic assigned

 

6. Graph-based Authentication Schemes to Support Blockchain-Based Identity Management Systems

The attractive properties of Distributed Ledger Technology (DLT) guarantee tamper resistant bookkeeping of data in a decentralized ledger without the need of a third party. However, not all blockchain applications can rely on closed loop verification of transaction input before the data appendage to the immutable ledger. Especially with consortium-based DLT, stakeholders and network components require fair control over verification of external input. In the context of blockchain-based identity management systems, bindings between digital identity objects enhance verification possibilities and increase the design space for tailored identity attribution. For this topic, you should should investigate how the authentication of membership can supports decisions in blockchain-based identity management procedures.

References: 

Supervisor: Jan Lauinger

Topic assigned

 

7. WiFi Authentication and Key Generation for IoT

This seminar topic investigates how client devices connect to WiFi hotspots and how encryption keys are generated using WPA2/WPA3. Your task is to research, understand, and summarize the working principles of the used protocols such as Open System Authentication and Four-Way Handshake. Furthermore provide a comparison of the different authentication methods and state which methods are implemented in typical WiFi-enabled microprocessors such as the ESP32.

References:

Supervisor: Emanuel Regnath

Topic assigned

 

8. Lattice-based Post-Quantum Signatures

Quantum computing poses challenges to public key signatures, because current schemes such as RSA and ECC will be broken by Shor's algorithm. NITS currently evaluates new proposals for quantum-secure signatures and one promising family are lattice-based signatures. Your task is to investigate and explain how lattice-based signatures work by providing a toy example using small numbers and/or visualization. A comparison of the different candidates against each other and in terms of performance against ECDSA should also be provided.

References:

Supervisor: Emanuel Regnath

Topic assigned

 

9. Performance Analysis Problems in TSN Networks

Performance analysis/schedulability analysis is another important topic in TSN networks. Formal performance analysis methods, such as network calculus and formal perfomance analysis, based on mathemathical model are used to give the upper bounds of worst-case delay and backlogs. Simulation model provides statistical results but not deterministic guarantees, by capturing the behaviour of the real network. Recently, machine learning is also used for assessing schedulability under some accuracy range. For this topic, you should learn about the performance analysis methods. You should give a comparison of current existing research on performance/schedulability analysis problems in TSN networks.

References:

Supervisor: Luxi Zhao

Topic assigned

 

10. Reconfiguration Problems in TSN Networks

Time-Sensitive Networking (TSN) is a set of Ethernet sub-standards defined in the IEEE 802.1 TSN Task Group, aiming at providing Ethernet with real-time communication, reconfiguration and reliability services in the layer 2 of network architecture. Some configuration (scheduling) problem for TSN has been addressed at the design time, however, there is a strong trend towards systems allowed dynamic reconfiguration at runtime to meet the new requirements, especially for automotive domain. TSN has high complexity of reconfiguration in case of mixed-criticality applications being used more and more, so currently available static approaches is time consuming. For this topic, you should learn about the flexible configuration framework and different scheduling mechanisms in TSN networks. You should give an overview of current existing research on reconfiguration problems in TSN networks.

References:

Supervisor: Luxi Zhao

Topic assigned

 

11. Runtime Monitoring / Verification for Cyber-Physical Systems

Verification usually refers to the offline task of checking, whether a system can or cannot violate a property during its execution. For large-scale, complex, or distributed systems, verifying a property can be infeasible, because the search space is too large to be exhaustively analyzed within a reasonable time. In this case, it can make sense to evaluate a property at runtime. For this topic, you should investigate the different approaches for runtime monitoring / verification for Cyber-Physical Systems.

References:

Supervisor: Laurin Prenzel

Topic assigned

 

12. Formal Specifications for Cyber-Physical Systems

To know if a system is running according to your needs, you need requirements. To automatically check, whether the system fulfills your requirements, you need formal requirements. Within Cyber-Physical Systems (CPS), trial-and-error methods are widespread and the use of formal methods is lacking. For this topic, you are expected to gather the requirements CPS pose on formal methods, get an overview of available formal specification languages, and evaluate the suitability of a selection of those languages for CPS.      

References:

Supervisor: Laurin Prenzel

Topic assigned

 

 

13. Using Artificial Intelligence Mechanisms to Implement Intrusion Response System

Safety-critical systems, such as autonomous vehicles, are often subject to various attacks. Within such systems, the response to an attack is as pivotal as the detection of the attack itself. Thus, the need for a proactive and distributed intrusion response system (IRS) becomes a critical requirement to ensure the secure and safe operation of these systems. Technologies like deep learning and reinforcement learning had shown good outcomes when they were adopted to detect the cyber-attacks. These technologies seem to be an excellent candidate to support the design and implementation of the IRS too. In the topic, you need to look at the recent research activities that investigated the use of artificial intelligence (AI) in the field of intrusion response. Also, you need to discuss their founds, present the open challenges, and propose some solutions for these objections.

References:

Supervisor: Mohammad Hamad

Topic assigned

To top