Persönlicher Status und Werkzeuge

Open Theses

You may also contact one of our doctoral candidates directly, if you are interested in a Bachelor or Master thesis, a student job, an "Ingenieurspraxis" or a "Forschungspraxis". It often happens, that topics are being prepared which are not adverdised yet. It may also be possible to find a topic matching your specific interests.
Please include a curriculum vitae together with a list of attended courses when applying for a thesis.
If your "Ingenieurspraxis" is selected to be supervised by one of our professors, please hand in the documents to Doris Dorn (Room N2401).

Bachelor Theses

Interference Coordination in a Local Area Scenario

The performance of current cellular networks is degraded by interference between neighboring cells (so called inter-cell interference). It can be improved if the base stations share the channel state information of the useful and interfering links. With this information the base stations can coordinate their transmission strategies. This is called interference coordination.

One form of interference coordination is large-scale MIMO. For many antennas at each base station it is possible to null the interference caused to other users, while serving the users associated to the base station.

The student should implement large-scale MIMO and compare it to network MIMO in a local area scenario.

Prerequisite: Linear Algebra, Digital Communications (Nachrichtentechnik)

Supervisor: Stefan Dierks

Master/Diploma Theses

A Deterministic Approach to the Gaussian Multiple-Input Multiple-Output (MIMO) Broadcast Channel
Multi-user communication systems (such as wireless systems) are different than point to point channels in two key aspects; namely, the broadcast nature of transmissions (e.g., wireless transmissions can be picked up by any receiver in the vicinity of the transmitter) and the superposition (e.g., a wireless receiver picks up a linear combination of transmissions from all transmitters in its vicinity). Recently a deterministic approach to multi-user information theory has proved promising by achieving progress in some of the long-standing open problems by finding approximate solutions. The idea behind a deterministic approach, on a high level, is to focus on the interaction of signals rather than the background noise, through a simpler appropriate model; find exact solutions for the simpler model; and then translate the intuitions and techniques to new techniques and/or approximate solutions on noisy models.
The goal of thesis is to undertake a deterministic approach to Gaussian MIMO broadcast channels.
Prerequisite: Linear Algebra, Information Theory
Supervisor: Shirin Saeedi

Role of Feedback in Broadcasting
This project intends to investigate the role of feedback in broadcast channels. More information available upon request. Pre-requisite (at least one of the following):
Basic knowledge of Network Coding (see [2])
Basic knowledge of Algorithms
Basic knowledge of Information Theory
Supervisor: Shirin Saeedi

Bayesian inference and tracking techniques for non-Lorentzian lineshape optical sources

Project description
Research on optical communication systems is nowadays focusing on methods and technologies which would allow higher capacity and increased transmission distance. The capacity of optical communication systems can be increased by a combination of spectrally efficient modulation formats and advanced receiver design. Currently, commercially available optical communication systems employ modulation formats that carry 2 bits/symbol/polarization, such as quadrature phase shift keying. In the near future, it is expected that in order to satisfy capacity demands, it will be required to increase the spectral efficiency to 4 or even 6 bits/symbol/polarization. This would require moving to 16 or 64 quadrature amplitude modulation (QAM).
Moreover, there is a strong focus on designing optical systems with reduced footprint and this requires large scale integration of optical components including semiconductors lasers. The semiconductor lasers exhibit a non-Lorentzian lineshape due to carrier induced frequency noise. This noise presents itself mainly as a resonance peak on the frequency noise power spectral density (PSD), which has been shown to play an important role on the system performance [4]. If phase noise tracking algorithms are not properly designed a large penalty in system performance may be obtained for higher order modulation formats (> 16 QAM) [4-5]. For non-Lorentizan lineshape more sophisticated tracking algorithms are therefore needed as the physics behind the laser phase noise process needs to be taken into account. As a proof-of-principle, it has been demonstrated theoretically and experimentally that Bayesian filtering methods are an effective tool in tracking phase noise that exhibits non-Lorentizan lineshape [5-6]. One of the challenges with Bayesian tracking methods is that they require knowledge of the parameters of the state-space describing the system. This is especially challenging to infer from the measured data due to the high dimensionality of the state-space as well as that the measured systems have typically a large number of unknown variables [7].

Objectives of the project
The main topic of this Master Project is implementation of Bayeisan inference techniques for parameter identification of state-space model for optical sources exhibiting non-Lorentzian lineshapes. The student will develop framework for parameter estimation based on expectation maximization which is an iterative approach for maximum likelihood parameter estimation. The inferred knowledge about the state-space will then be used to formulate digital signal processing algorithms for carrier phase noise compensation for coherent optical communication systems. If time allows, the student will test the framework on experimental data obtained for an optical transmission systems employing long haul transmission and higher order model formats.

Relevant literature
[1] Govind Agrawal, Fibre-optic communication systems, 2010
[2] Dirk van der Borne, PhD Thesis, 2008
[3] Essiambre et al., Journal of Lightwave Technology, vol. 28, 2010
[4] M. Piels, in Proceedings of OFC 2015
[5] L. Barletta, Journal of Lightwave Technology, vol. 30, no. 10, 2012
[6] D. Zibar, Journal of Lightwave Technology, vol. 33, no. 7, 2015
[7] C M. Bishop, Pattern recognition and machine learning, 2006

Supervisor: Luca Barletta, Darko Zibar (DTU)

Modulation and Channel Coding

Scope:I offer master thesis topics in the area of modulation, signaling and channel coding. The goal is to design transmission schemes for reliable communication close to the theoretical limits. We consider wireless, optical and wired communication channels and take power, bandwidth, and latency constraints into account.

Prerequisites: you should have attended one or more of the following courses: Information Theory, Channel Coding, Coded Modulation, Channel Codes for Iterative Decoding.

Start: October 1st.

Supervisor: Georg Böcherer

Information Theory
I offer bachelor and master thesis topics in the area of information theory. The goal is to understand the fundamental limits of communications systems. Please contact me if you are interested.
Supervisor: Roy Timo

Probabilistic Signal Shaping for Short-Reach Optical Fiber Systems
Optical fiber systems form the backbone of the internet. Despite their large usable bandwidth, they are going to running out of capacity due to an increasing demand for services such as high-definition video streaming. In order to obtain the largest possible transmission rates for a fixed link, a non-uniformly spaced input can replace the common uniform input. Prerequisites for this thesis: digital communications (must!), basics in optical communication systems (important), information theory (equally important). If you are interested in this work, please contact me at to set up a personal meeting.
Supervisor: Tobias Fehenberger

Efficient Algebraic Decoding of Subspace Codes for Network Coding
Subspace codes were proposed by Kötter and Kschischang in 2008 for error control in random linear network coding (RLNC). The main idea is to encode information by the choice of a subspace rather than a choice of a vector. A basis for the selected subspace is then transmitted over a channel. The decoder returns the subspace that is closest to the received subspace according to some metric.
I offer topics on efficient (list and unique) decoding algorithms for subspace and rank-metric codes that allow to correct errors beyond the unique decoding radius.

You should be interested in Channel Coding and should have good knowledge in (linear) Algebra.
Supervisor: Hannes Bartz

Rate-Distortion Theory for Sparse Sources
Rate-Distortion Theory for Sparse Sources This project investigates information theory limits for the lossy compression of sparse sources. Motivated by the recent mathematical theory Compressed Sensing that deals with the efficient acquisition and reconstruction of sparse signals, we try to determine the fundamental limits of digitally storing data created by such sources. A sparse source emits a string of symbols of which - loosely speaking - only a few contain information. There are different ways to model these sources. An example is the Gaussian spike source which emits a Gaussian symbol with probability p and a zero with probability 1-p.

Possible directions of a project are:
- Investigating different source models
- (Numerically) determining the rate distortion function of a sparse source
- Computing finite length performance bounds for sparse sources

- Information Theory (must)
- Pleasure with mathematics (must)
- Multiuser information theory (beneficial)
Supervisor: Lars Palzer

Open Student Jobs

no news in this list.

Theses in Progress

Bachelor Theses

Dina Khaled Sayed Abdel-Hadi: Information Theory of Multi-mode Optical Fiber
Space-division multiplexing (SDM) over multi-mode fiber has recently been suggested as a solution to meet the ever-increasing demand for higher data rates. This poses the problem of finding bounds on the capacity of multi-mode fiber to quantify the possible gains of applying SDM. The answer to the question of capacity of fiber lies at the intersection of fiber modeling with information theory. Capacity estimates depend on the model of the channel under consideration. As a result, a thorough understanding of the propagation of waves through fiber is needed to model the channel adequately. Then, the tools of information theory are applied to such models to find the capacity. In this thesis, a model of the fiber channel that takes into account several linear and nonlinear effects is derived. Then, the capacity is estimated for a simplified model with linear coupling and white Gaussian noise.
Supervisor: Mansoor Yousefi

Stephanie Dangl: A Soft-Information Model for Random Linear Network Coding
Random Linear Network Coding (RLNC) is a powerful approach to achieve the capacity in multicast networks. Subspace codes are an efficient family of error correction codes in RLNC. The idea behind these codes is to map information on a subspace and to transmit a basis for this subspace. In this thesis, a model incorporating soft-information (reliability information) of the received packets (basis vectors) using a linear inner code is considered. First the student should become familiar with linear block codes (e.g BCH, Reed--Solomon codes), concatenated coding schemes and the basics of Generalized Minimum Distance (GMD) decoding. In the next step, suitable inner codes to acquire the reliability information of the received packets should be chosen and implemented in MATLAB. The performance of the coding schemes should be evaluated and compared by performing simulations in MATLAB.
Supervisor: Hannes Bartz, Vladimir Sidorenko

Achraf Kamoun: Balance Problems, Information Theory, and PPM Demodulation
Balance problems are beautiful mathematical puzzles, an instance of which goes like that: "There are nine balls, one of which is heavier. You have a balance scale. How many weighings are necessary to determine the heavy ball?" To this particular instance, the answer is two [1]. That two weighings are not only sufficient, but also necessary (i.e., two weighings are optimal), can be shown using information theory: The uncertainty about the heavy ball is equal to the information that is obtained with two weighings. There are several variations on this problem: Arbitrary numbers of balls can be considered, or it may be unknown whether the oddball is heavier or lighter than the rest. We believe that information theory can tell us something about these variations, too. In this thesis, these variations will be analyzed using standard tools from information theory: conditional entropy and the chain rule. Finally, the connection to non-coherent demodulation of pulse-position-modulated signals will be investigated.

[1]: Blog post
Supervisor: Bernhard Geiger

Tobit Klug: Modelling and Mitigation of impulsive noise in PLC transmission system
Supervisor: Elisabeth Oberleithner

Andrea Nicholas Beretta: Funktionale Absicherung der Informations- und Kommunikationssysteme (I&K) im Automotive Umfeld für verschiedene Funktionen
Supervisor: Gerhard Kramer, Fa. Telemotive AG, Frankfurter Ring 115a. 80807 München

Jun Jian Tay: Performance Analysis of Regular and Spatially Coupled LDPC Codes
Spatially coupled (SC-) LDPC codes are known to reach capacity. Yet, it is unknown what the advantages in the finite-length regime are. The task of the student is to compare coupled LDPC codes with their underlying counterpart for finite code lengths using finite-length scaling laws and to verify the results by simulations.
Supervisor: Markus Stinner

Master/Diploma Theses

Stefan Thiel: Analysis of AM Demodulation for Speech-based Airborne Communications Under Multipath Propagation
This master thesis is in cooperation with the radio communications department at Rohde&Schwarz. Airborne radio communication is based on Amplitude Modulation (AM). To use them with existing low-directivity antennas in airplanes, the effects of multipath propagation have to be considered. The student will investigate the channel effects on airborne communications (Ground-Air, Air-Air) and compare possible receiver designs. For this aim, realistic multipath-propagation scenarios will be defined for airborne communications. The models and the suggested designs will be tested with Matlab/Simulink. The student will evaluate the results and compare multiple techniques to improve communication quality.
Supervisor: Onur Günlü, Dipl.-Ing. Ludger Korsmeier and Dipl.-Ing. Michael Gerth (Rohde&Schwarz)

Anastasios Dimas: Peeling Decoding Analysis of Non-binary LDPC Codes
The performance of non-binary LDPC (NB-LDPC) codes on the binary erasure channel (BEC) is analyzed. Existing analysis tools of their binary counterparts are extended for NB-LDPC codes. Therefore, the binary extension of the NB-LDPC codes is constructed and its behaviour is investigated analytically.
Supervisor: Markus Stinner

Arun Sattanathan Gothandaraman : Firmware development and testing for Optical communication transceivers
Tasks & Responsibilities: - Architect, design, code, debug, test and document new/existing designs with a major focus on the embedded hardware elements of the design - Work to minimize the need for third party development resources bringing into balance in-house vs. outsourced services - Collaborate with HW designers on the selection of components for new designs where applicable - Develop test procedures and support design of test and design qualification systems- - Provide documentation of all designs
Supervisor: Norbert Hanik, Julio Ramirez (DLR)

Frank Löw: Identification, Development, and Simulation of Protocol Concepts for a Battery-Driven Backbone-Router
The aim of this thesis is to develop low-power protocols for a multi-hop backbone network in a wireless sensor network.
Supervisor: Youlong Wu, Shirin Saeedi, Hekatron Vetriebs GmbH, Philipp Alber

Boxiao Ma: Multipath Assisted Positioning: Mapping of Virtual Transmitters
Positioning is one of the most important fields of application for wireless radio transmissions. In critical environments, such as urban canyon or indoors, the position accuracy using wireless signals is drastically reduced. In these environments, multipath effects, low received signal power and non-line-of-sight propagation reduce the position accuracy. With Channel-SLAM, we proposed a paradigm shift in how to process the received signal in order to provide accurate position estimation for mobile receivers: rather than mitigating multipath, we propose to exploit multipath. Channel-SLAM treats multipath components as signals emitted from virtual transmitters which are time synchronized to the physical transmitter and static in their position. Hence, multipath propagation increases the number of transmitter by virtual transmitters resulting in more accurate position estimation or enabling positioning in situations where the number of physical transmitters is insufficient. Channel-SLAM estimates the receiver position and the positions of the virtual transmitters simultaneously and does not require any prior information such as room-layout or a database for fingerprinting. This Master thesis builds on and extends the previous work on Channel-SLAM. In order to exploit multipath components for multiple receivers, this Master thesis shall derive a mapping algorithm for virtual transmitters. The positions of virtual transmitters do not change over time, however, the virtual transmitters are only visible in a certain area. Hence, each virtual transmitter can be described by a static position and a visibility region. Thus, this master thesis shall derive a representation of the visibility region of virtual transmitters and shall extend the current Channel-SLAM approach. Additionally, the results should be verified by simulations and measurements. The results might give valuable geometric information on the locations of reflector and scatterers, which might enable to obtain geometric information of all kinds of environments by using wireless signals.
Supervisor: Norbert Hanik, Christian Gentner (DLR)

Yuchen Wu: Higher-Order Markov Aggregations
In [1], we used an information-theoretic characterization of lumpability as a cost function for Markov chain aggregation, i.e., for replacing a Markov chain on a large state space by a Markov chain on a smaller state space. We showed that the cost function can be interpreted as the Kullback-Leibler divergence rate between the original chain and chain that is obtained by lifting the aggregated one. We furthermore showed that by relaxing the problem, the information bottleneck method can be used for obtaining the aggregation.

This thesis should extend the work to higher-order aggregations, i.e., to replacing a Markov chain on a large state space by a higher-order Markov chain on a smaller state space. Higher-order Markov chains allow for smaller modelling errors but come at the cost of a higher model complexity -- investigating higher-order aggregations thus admit trading one design goal for the other.

The main questions are whether the corresponding cost function can also be interpreted as a Kullback-Leibler divergence rate and whether the information bottleneck method can again be employed for aggregation. In case we find a lifting that admits an interpretation as a Kullback-Leibler divergence rate, it is worth changing the lifting method to the one presented in [2], investigating connections to the spectral theory of Markov chains.

[1]: B. C. Geiger, T. Petrov, G. Kubin, and H. Koeppl, “Optimal Kullback-Leibler aggregation via information bottleneck,” IEEE Trans. Autom. Control, vol. 60, no. 4, pp. 1010–1022, Apr. 2015, open-access: arXiv:1304.6603 [cs.SY].
[2]: K. Deng, P. G. Mehta, and S. P. Meyn, “Optimal Kullback-Leibler aggregation via spectral theory of Markov chains,” IEEE Trans. Autom. Control, vol. 56, no. 12, pp. 2793–2808, Dec. 2011.
Supervisor: Bernhard Geiger

Nicolas Blum: Algorithm Development for FFT-based FM Multitone Signal
Supervisor: Andrei Nedelcu, Joerg Biedermann (Rohde & Schwarz)

Bogdan Balauta: Code Design for Secrecy

The thesis consists of finite-length code design for security primitives for given channel parameters, desired block lengths and BERs.

Prerequisites: Channel Coding lecture is necessary, any other advanced channel coding lecture is a plus.

Supervisor: Onur Günlü


Dominik Fienko: Integration BS2000 SE Serie in Enterprise System Management
Supervisor: Gerhard Kramer, Fa. Fujitsu Technology Solutions, PDG ES S SWE 057, München

Mahmoud Marcam Helmy: Erstellung einer App für Smartphones der BITS mobile ERP Linie auf Basis von IOS Plattformen
Supervisor: Gerhard Kramer, BITS better it solutions GmbH, Wallenmahd 47, A-6850 Dornbirn

Mohamed Salah Ben Slimen : SDR - Digital Modulation Real Time Receiver in Simulink

Software Defined Radio is a relatively new concept in thinking and designing communications systems. What defines it is that almost all parts of a transciever (transmitter/receiver) are implemented as software modules on a general purpose computer, rather than on dedicated hardware. The radio frequency part of the receiver is thus also reconfigurable through software and will represent just an interface.

Simulink is a graphical programming tool from Mathworks and is especially good for fast prototyping and simulating of complex systems. It provides many custom made software modules and light and easy way to set up graphical user interfaces.

The purpose of this project is that following the principles of SDR radio and the facilities offered by Matlab/Simulink you could build a real-time receiver for digital modulated transmissions. Links:

Supervisor: Andrei Nedelcu

Ulrich Patrick Sandoval: Weiterentwicklung des Netzwerkanalysetools (NAT) mit Schwerpunkt auf Testfälle
Supervisor: Gerhard Kramer, Fa. Infineon Technologies, Neubiberg, Hr. Lippmann

Lei Zhang: Optimierung und Qualitätssicherung des Vodafone GSM-, UMTS- und LTE-Mobilfunknetzes
Supervisor: Norbert Hanik, Fa. Vodafon e GmbH, München, Hr. Quicel

Hasan Yagiz Özkan: Image Tracking Algorithm Development
Supervisor: Gerhard Kramer, Fa. Bases Teknoloji Arge Sanay, Hr. T. Cirak, Istanbul

Burak Tas: Analog- und Digital-Radio: Konzeption und Automatisierung von Testspezifikationen
Supervisor: Gerhard Kramer, Telemotive AG, Breitwiesen, 73347 Mühlhausen

Simon Heine: Softwareentwicklung für Hochfrequenz- und taktische Funkgeräte
Supervisor: Gerhard Kramer, Fa. Rohde & Schwarz GmbH & Co. KG, Mühldorfstr. 15, 81671 München

Jonas Zucker: Kalibrierung/Justage der Ipetronik internen Datenerfassungsmodule
Supervisor: Norbert Hanik, IPETRONIK Eichstätt GmbH, Industriestr. 10, 85072 Eichstätt, Hr. S. Löffler

Ahmet Burakhan Koyuncu: Hardwarenahe Optimierung und Laufzeitverbesserungen der Software, Design und Optimierung von Schnittstellen
Supervisor: Norbert Hanik, MVTec Software GmbH, Neherstr. 1, 81675 München, Hr. T. Hopfner

Tim Heinlein: Einarbeitung in das automatisierte Regression Test System der Data Application Unit
Supervisor: Norbert Hanik, Fa. Rohde & Schwarz GmbH & Co. KG, 81671 München, Hr. T. Rösner

Muhammed Efe: Improving lab experiment for Pulse Coded Modulation
The student will take care of following tasks:

1. Modify the Matlab code for PCM lab experiment to make it compatible with the current audio interface.
2. Modernize the Graphical User Interface for the experiment.
3. Improve the description and the assignments for the experiment.
Supervisor: Gerhard Kramer, Ali Amjad, Markus Stinner

Lukas Püttner: Evaluierung verschiedener Tools zur Extraktion und Auswertung von Metadaten aus IP-Verkehrsdaten
Supervisor: Norbert Hanik, ESG Elektroniksystem- und Logistik GmbH, 82256 Fürstenfeldbruck

Christian Bogner: Identifizierung und Verfolgung der Out-of-Spec Geräte in Hinblick auf ISO9001 Compliance
Supervisor: Norbert Hanik, Fa. Intel Deuschland

Mohamed Chennoufi: Website/wiki for Code performance
Aim of the project is to build a basic platform where users can upload error correction codes that are evaluated and results are shown on the wiki.
Supervisor: Patrick Schulte

Francesco Montanari: Dynamic Dead Pixel Correction
Supervisor: Norbert Hanik, Fa. ARRI, Arnold & Richter Cine Technik GmbH & Co KG, München, Fr. E. Kögel

Forschungspraxis (12 ECTS)

Mustafa Cemil Coşkun: Staircase Codes for Optical Communications
Staircase codes are a relatively new class of algebraic channel codes for high-speed optical communications. Their large coding gains and reasonable decoding complexity have made them a promising option for data rates of 100 Gbit/s and more. The student's task of this Forschungspraxis is to understand the concept of staircase codes and implement them in Matlab or C.
Supervisor: Tobias Fehenberger, Georg Böcherer

Yunhong Ju: Model Selection for real data
The student will first learn how to apply different model selection methods. The second step is the comparison of these methods in terms of their performance for different source statistics and sample sizes. The selection of the best modeling method and the best model for real security primitive output data will follow. The student is expected to write a report about the comparison results of selected methods and implement a Matlab script illustrating the results.
Supervisor: Onur Günlü

MSCE Internships

Dinesh Manobar Gandhinathan: Maximally Redundant Communication Channels
Consider the following problem: Let X and Y have the same alphabet and a given joint distribution giving rise to mutual information I(X;Y). Let W be a discrete, memoryless channel, and define U (V) as the output of W with X (Y) at the input. By the data processing inequality, for all W we will have that I(U;V) is at most I(X;Y). The question is now, how one should choose W in order to maximize the mutual information I(U;V).

If the channel output alphabet is at least as large as the alphabet of X and Y, the solution is trivial. Moreover, if the output alphabet is smaller, and if for X and Y different channels W1 and W2 can be used, the optimal channel pair is deterministic, i.e., both W1 and W2 are described by non-injective functions. It can be shown that this simplification is not possible if we require W1=W2: The optimal channel W is, in general, stochastic.

In this internship, the properties of the mutual information I(U;V) as a function of the channel W should be discussed. Based on these properties, an algorithm should be designed that maximizes this mutual information over the set of stochastic matrices with given dimensions. The results have implications for soft clustering of data given by pairwise distances and for the aggregation of Markov chains.
Supervisor: Bernhard Geiger