Sperm-driven micromotors, consisting of a single sperm cell captured in a microcap, utilize the strong propulsion generated by the flagellar beat of motile spermatozoa for locomotion. It enables the movement of such micromotors in biological media, while being steered remotely by means of an external magnetic field. The substantial decrease in swimming speed, caused by the additional hydrodynamic load of the microcap, limits the applicability of sperm-based micromotors. Therefore, to improve the performance of such micromotors, we first investigate the effects of additional cargo on the flagellar beat of spermatozoa. We designed two different kinds of microcaps, which each result in different load responses of the flagellar beat. As an additional design feature, we constrain rotational degrees of freedom of the cell’s motion by modifying the inner cavity of the cap. Particularly, cell rolling is substantially reduced by tightly locking the sperm head inside the microcap. Likewise, cell yawing is decreased by aligning the micromotors under an external static magnetic field. The observed differences in swimming speed of different micromotors are not so much a direct consequence of hydrodynamic effects, but rather stem from changes in flagellar bending waves, hence are an indirect effect. Our work serves as proof-of-principle that the optimal design of microcaps is key for the development of efficient sperm-driven micromotors.

Cellular micromotors are attractive for locally delivering high concentrations of drug, and targeting hard-to-reach disease sites such as cervical cancer and early ovarian cancer lesions by non-invasive means. Spermatozoa are highly efficient micromotors perfectly adapted to traveling up the female reproductive system. Indeed, bovine sperm-based micromotors have shown potential to carry drugs toward gynecological cancers. However, due to major differences in the molecular make-up of bovine and human sperm, a key translational bottleneck for bringing this technology closer to the clinic is to transfer this concept to human material. Here, we successfully load human sperm with Doxorubicin (DOX) and perform treatment of 3D cervical cancer and patient-representative ovarian cancer cell cultures, resulting in strong anticancer cell effects. Additionally, we define the subcellular localization of the chemotherapeutic drug within human sperm, using high-resolution optical microscopy. We also assess drug effects on sperm motility and viability over time, employing sperm samples from healthy donors as well as assisted reproduction patients. Finally, we demonstrate guidance and release of human drug-loaded sperm onto cancer tissues using magnetic microcaps, and show the sperm microcap loaded with a second anticancer drug, camptothecin (CPT), which unlike DOX is not suitable for directly loading into sperm due to its hydrophobic nature. This co-drug delivery approach opens up novel targeted combinatorial drug therapies for future applications.

The millimeter wave (mmW) band, commonly referred to as the frequency band between 30 GHz and 300 GHz, is seen as a possible candidate to increase achievable rates for mobile applications due to the existence of free spectrum. However, the high path loss necessitates the use of highly directional antennas. Furthermore, impairments and power constraints make it difficult to provide full digital beamforming systems. In this thesis, we approach this problem by proposing effective beam alignment and beam tracking algorithms for low-complex analog beamforming (ABF) systems, showing their applicability by experimental demonstration. After taking a closer look at particular features of the mmW channel properties and introducing the beamforming as a spatial filter, we begin our investigations with the application of detection theory for the non-convex beam alignment problem. Based on an M-ary hypothesis test, we derive algorithms for defining the length of the training signal efficiently. Using the concept of black-box optimization algorithms, which allow optimization of non-convex algorithms, we propose a beam alignment algorithm for codebook-based ABF based systems, which is shown to reduce the training overhead significantly. As a low-complex alternative, we propose a two-staged gradient-based beam alignment algorithm that uses convex optimization strategies after finding a subregion of the beam alignment function in which the function can be regarded convex. This algorithm is implemented in a real-time prototype system and shows its superiority over the exhaustive search approach in simulations and experiments. Finally, we propose a beam tracking algorithm for supporting mobility. Experiments and comparisons with a ray-tracing channel model show that it can be used efficiently in line of sight (LoS) and non line of sight (NLoS) scenarios for walking-speed movements.

To enable a new level of connectivity among machines as well as between people and machines, future wireless applications will demand higher requirements on data rates, response time, and reliability from the communication system. This will lead to a different system design, comprising a wide range of deployment scenarios. One important aspect is the evolution of physical layer (PHY), specifically the waveform modulation. The novel generalized frequency division multiplexing (GFDM) technique is a prominent proposal for a flexible block filtered multicarrier modulation. This thesis introduces an advanced GFDM concept that enables the emulation of other prominent waveform candidates in scenarios where they perform best. Hence, a unique modulation framework is presented that is capable of addressing a wide range of scenarios and to upgrade the PHY for 5G networks. In particular, for a subset of system parameters of the modulation framework, the problem of symbol time offset (STO) and carrier frequency offset (CFO) estimation is investigated and synchronization approaches, which can operate in burst and continuous transmissions, are designed. The first part of this work presents the modulation principles of prominent 5G candidate waveforms and then focuses on the GFDM basic and advanced attributes. The GFDM concept is extended towards the use of OQAM, introducing the novel frequency-shift OQAM-GFDM, and a new low complexity model based on signal processing carried out in the time domain. A new prototype filter proposal highlights the benefits obtained in terms of a reduced out-of-band (OOB) radiation and more attractive hardware implementation cost. With proper parameterization of the advanced GFDM, the achieved gains are applicable to other filtered OFDM waveforms. In the second part, a search approach for estimating STO and CFO in GFDM is evaluated. A self-interference metric is proposed to quantify the effective SNR penalty caused by the residual time and frequency misalignment or intrinsic inter-symbol interference (ISI) and inter-carrier interference (ICI) for arbitrary pulse shape design in GFDM. In particular, the ICI can be used as a non-data aided approach for frequency estimation. Then, GFDM training sequences, defined either as an isolated preamble or embedded as a midamble or pseudo-circular pre/post-amble, are designed. Simulations show better OOB emission and good estimation results, either comparable or superior, to state-of-the-art OFDM system in wireless channels.

This work presents a distance measurement system for the localization of passengers in the sea in case of a ship sinking emergency. By applying the radar principle together with a switched injection-locked oscillator (SILO) as an active reflector the distance to a passenger can be determined through continuous measurement of the roundtrip time of flight of a chirp signal. The system is based on a concept previously published in [Wie03, VG08, Str14] and for the first time is designed in the 2,45GHz ISM frequency band. Criteria for the design of SILOs formulated in [Str14] were applied and verified for the first time at an operating frequency of 2,45GHz. The designed circuits were manufactured in a SiGe BiCMOS technology. The SILOs are based on a commonbase Colpitts and cross-coupled topology and exhibit the lowest published input referred noise power of −79 dBm. The output power is 12,3 dBm and 12,6 dBm, respectively. The efficiency of the cross-coupled as well as Common-Base Colpitts SILO is 26%. The theory of a SILO was extended towards the influence of interfering signals and the modulation signal. An interfering signal influences the starting behavior of a SILO in a way that a distance measurement leads to incorrect results. It was shown that interfering signals in the ISM frequency bands of 868MHz, 2,4GHz and 5,8GHz don’t allow correct distance measurements due to their high output power. Therefore, it is recommended to use an operating frequency of the SILO that is not close to any possible interfering signal, like the ISM band at 24GHz. This minimizes measurement errors and enables a more accurate distance measurement. Furthermore, it was shown that a phase coherent start of oscillation to any received signal is influenced by the modulation signal. For injection powers below −40 dBm the oscillator does not start its oscillation due to the injected signal, but to the modulation signal itself. This disables exact distance measurements. Through the use of a modulation capacitor at the base of the tail current source this effect can be minimized and the input referred noise power of the oscillator can be improved. The functionality of the distance measurement system was also verified for both SILO topologies. The range of the system in the best case scenario with a Common-Base Colpitts SILO was 120m, at an accuracy of 53 cm and a precision of 42 cm. All results regarding accuracy and precision exceed the specification of the system.

Computer engineering is advancing rapidly. For 55 years, the performance of integrated circuits has almost doubled every 18 months. Mostly, these advancements were enabled by technological progress. Even the end of frequency scaling could not bring the ever-increasing performance growth to a halt. However, technology burdens, like noticeable leakage currents, have piled up, which shifts the focus towards architectural improvements. Especially the multi-core paradigm has proven its virtue for chip designs over the last decade. While having been introduced in high-performance computing areas, modern technology nodes also enable low-cost, low-power embedded designs to benefi t from multiple cores and accelerators. Since the majority of cores depend on memory, which requires a considerable amount of chip area, this common resource needs to be shared effi ciently. High-performance cores use shared caches to increase memory utilization. However, many accelerators do not use caches as they need predictable and fast scratchpad memory (SM). But sharing SM entails confl icts, questioning its fast and predictable nature. Hence, the question arises on how to adapt architectures for sharing while retaining SM’s advantages. This thesis presents a novel, shared SM architecture that embraces the idea of a minimal logic path between core and memory, thereby increasing the maximum operating frequency. Because of its additional capabilities, like dynamic address translation and programmable priorities, it is also well suited for heterogeneous platforms that use dynamic scheduling and require predictable behavior. Demonstrating its advantages, we analyze the characteristics of the new architecture and compare it to state-of-the-art approaches. To further mitigate confl icts, we present the conception of access interval prediction (AIP). By predicting memory accesses with a granularity of a single clock cycle, AIP guides the allocation of resources. This method maximizes memory utilization while reducing confl ict delays. With the help of various methods inspired by branch prediction, we achieve over 90 % of accurate predictions and reduce stall cycles signifi cantly. Another key contribution of this thesis is the extension of analytic models to estimate the throughput of shared SM systems. Again, the focus lies on heterogeneous systems with different priorities and access patterns. The results show a promising error reduction, boosting the used models applicability for real design use cases.

Das Thema der Arbeit ist ein Roboter der zur Unterstützung von Pflegekräften eingesetzt werden soll. Der Roboter soll dabei auch in der Lage sein mit pflegebedürftigen Menschen zu kommunizieren und sich ihnen zu nähern. Dabei sollte der Roboter sich sozial verträglich verhalten. Dazu wird in dieser Arbeit ein Modul entwickelt, welches Anhand der Person und der Umgebung eine sozial verträgliche Position liefert, um mit der Person wie gewünscht zu interagieren. / The subject of the thesis is a robot that is to be used to support nursing staff. The robot should also be able to communicate with people in need of care and to approach them. The robot should behave in a socially acceptable way. For this purpose, a module is developed in this thesis, which provides a socially acceptable position based on the person and the environment in order to interact with the person as desired.

Der in dieser Arbeit verfolgte Ansatz ermöglicht die generelle Herstellung taktiler Displays mit fotolithographischen Methoden. Dabei werden alle Aktorelemente Schicht für Schicht und gleichzeitig während des Herstellungsprozesses erzeugt. Somit ist eine Miniaturisierung und Skalierung möglich, wie sie für Prozesse in der Halbleitertechnik üblich ist. Um die Aktorgröße jedes einzelnen Aktors zu kontrollieren, wurde das temperatursensitive Hydrogel Poly(N-Isopropylacrylamid) verwendet und zur Steuerung eine elektronikkompatible thermische Schnittstelle entwickelt. Dabei entsteht ein hochaufgelöstes thermisches Feld, welches aktorgenau den Zustand von jedem visuellen Pixel und taktilem Taxel einstellt. Es wurde ein System zur Wiedergabe von sowohl glatten als auch rauen Oberflachen und ein weiteres System zur Wiedergabe von Konturen aufgebaut. Da die Aktoren ihre optischen Eigenschaften verändern, haben die Displays damit zusätzlich auch eine visuelle, monochrome Funktionalität.:1 Einleitung 2 Präzisierte Problemstellung 3 Stand der Technik 3.1 Haptik 3.2 Kutane Displays 3.3 Taktile Wahrnehmung als Grundlage kutaner Displays 3.3.1 Frequenzabhängigkeit der Hautwahrnehmung 3.3.2 Sensortypen der Haut 3.3.3 Zusammengesetzte Wahrnehmung und Wahrnehmbarkeitsschwellen 3.3.4 Auswahl des Aktorfunktionsprinzips 3.4 Grundlagen stimulisensitiver Hydrogele 3.4.1 Arten von Gelen 3.4.2 Thermodynamisches Gleichgewicht der Gelquellung 3.4.3 Quellkinetik: Zeitliche Beschreibung der Quellung 3.4.4 Herstellung von Hydrogelstrukturen 3.5 Erzeugung eines Temperaturprofils zur Aktorsteuerung 4 Herstellung des Aktorchips und Elementansteuerung 4.1 Trägersubstratherstellung 4.1.1 Trägersubstrat-Schichtaufbau: vereinfachtes Modell 4.1.2 Projektionsschicht 4.1.3 Herstellung einer transparenten Schicht definierter thermischer Eigenschaften 4.2 Aktorarrays auf Hydrogelbasis 4.3 Fotolithografischer Prozess 4.3.1 Belichtungseinheit 4.3.2 Lichtleistung und Uniformität 4.3.3 Fotolithografische Maske 4.3.4 Proximity-Belichtung – 1:1 Schattenprojektion 4.4 Hydrogelsynthese 4.4.1 Vernetzergehalt 4.4.2 Fotoinitiator 4.4.3 Losungsmittel- und Monomerkonzentration 4.4.4 Lösungsmittelqualitat und Temperatur 4.4.5 UV-Intensitat 4.4.6 Lithografische Hydrogelstrukturierung 4.4.7 UV-Belichtungszeit 4.4.8 Belichtungsprozess und Lichtleistung 4.4.9 Auswirkungen von Sauerstoff auf die Hydrogelstruktur 4.5 Strukturierungsgrenze 4.6 Haftvermittlung zwischen Hydrogelaktoren und Tragersubstrat 4.7 Thermische Aktorsteuerung 4.7.1 Temperatursteuerung im Quellungsgleichgewicht 4.7.2 Quellverhalten bei schneller Temperaturanderung und kurzer Quellzeit 4.8 Fertigungstechnologie fur Grosflächenmikrostrukturierung 4.8.1 Herstellung von Grosflächenmaster 4.8.2 Elastische Displayabdeckung 4.8.3 Mikrofluidische Quellmittelversorgung des Aktorchips 5 Optoelektrothermischer Controller 5.1 Konzept Warmesteuerung durch Licht 5.2 Aktorchip-Aufnahmeeinheit mit Licht-Wärme-Umwandlung 5.3 Mechanischer und optischer Aufbau 5.4 Zusatzoptik für Projektion auf 2,5 Zoll Displaydiagonale 5.5 Umlenkspiegel 5.6 Leistungsparameter der verwendeten Projektions-Systeme 5.7 Projektorcharakteristik 5.8 Setup der Thermografiekamera 5.9 Optimierung und Charakterisierung der Aktorsteuerung 5.9.1 Design-, Betriebs- und Messgrosen der thermischen Steuerung 5.9.2 Dimensionierung und Optimierung der thermischen Steuerung 5.9.3 Zeitliche und räumliche Auflösung der Aktorsteuerung 5.9.4 Optische Temperaturbestimmung durch Kontrastfunktion 6 Multimodales Display 6.1 Displayfunktionen 6.1.1 Basisaktorchip als Aktorarray mit Substrat und Abdeckung 6.1.2 Monochrome Displayfunktion 6.1.3 Ausgabe taktiler Oberflächeneigenschaften 6.1.4 Taktile Kantenhervorhebung der Kontur 6.2 Kombinierte Monochrom und Displacementfunktionalität 7 Zusammenfassung 8 Ausblick

Der Workshop Methoden und Beschreibungssprachen zur Modellierung und Verifikation von Schaltungen und Systemen (MBMV 2015) findet nun schon zum 18. mal statt. Ausrichter sind in diesem Jahr die Professur Schaltkreis- und Systementwurf der Technischen Universität Chemnitz und das Steinbeis-Forschungszentrum Systementwurf und Test. Der Workshop hat es sich zum Ziel gesetzt, neueste Trends, Ergebnisse und aktuelle Probleme auf dem Gebiet der Methoden zur Modellierung und Verifikation sowie der Beschreibungssprachen digitaler, analoger und Mixed-Signal-Schaltungen zu diskutieren. Er soll somit ein Forum zum Ideenaustausch sein. Weiterhin bietet der Workshop eine Plattform für den Austausch zwischen Forschung und Industrie sowie zur Pflege bestehender und zur Knüpfung neuer Kontakte. Jungen Wissenschaftlern erlaubt er, ihre Ideen und Ansätze einem breiten Publikum aus Wissenschaft und Wirtschaft zu präsentieren und im Rahmen der Veranstaltung auch fundiert zu diskutieren. Sein langjähriges Bestehen hat ihn zu einer festen Größe in vielen Veranstaltungskalendern gemacht. Traditionell sind auch die Treffen der ITGFachgruppen an den Workshop angegliedert. In diesem Jahr nutzen zwei im Rahmen der InnoProfile-Transfer-Initiative durch das Bundesministerium für Bildung und Forschung geförderte Projekte den Workshop, um in zwei eigenen Tracks ihre Forschungsergebnisse einem breiten Publikum zu präsentieren. Vertreter der Projekte Generische Plattform für Systemzuverlässigkeit und Verifikation (GPZV) und GINKO - Generische Infrastruktur zur nahtlosen energetischen Kopplung von Elektrofahrzeugen stellen Teile ihrer gegenwärtigen Arbeiten vor. Dies bereichert denWorkshop durch zusätzliche Themenschwerpunkte und bietet eine wertvolle Ergänzung zu den Beiträgen der Autoren. [... aus dem Vorwort]

In der vorliegenden Arbeit wird das bestärkende Lernen im Kontext der Steuerung und Regelung nichtlinearer dynamischer Systeme untersucht. Es werden zunächst die Grundlagen der stochastischen Optimalsteuerung sowie des maschinellen Lernens, die für die Betrachtungen dieser Arbeit relevant sind, erläutert. Anschließend werden die Methoden des bestärkenden Lernens im Kontext der datenbasierten Steuerung und Regelung dargelegt, um anschließend auf drei Methoden des tiefen bestärkenden Lernens näher einzugehen. Der Algorithmus Deep-Deterministic-Policy-Gradient (DDPG) wird zum Gegenstand intensiver Untersuchungen an verschiedenen mechanischen Beispielsystemen. Weiterhin erfolgt der Vergleich mit einem klassischen Ansatz, bei dem die zu bewältigenden Steuerungsaufgaben mit einer modellbasierten Trajektorienberechnung, die auf dem iterativen linear-quadratischen Regler (iLQR) basiert, gelöst werden. Mit dem iLQR können zwar alle Steuerungsaufgaben erfolgreich bewältigt werden, aber für neue Anfangswerte muss das Problem erneut gelöst werden. Bei DDPG hingegen wird ein Regler erlernt, der das zu steuernde dynamische System – aus nahezu beliebigen Anfangswerten – in den gewünschten Zustand überführt. Nachteilig ist jedoch, dass der Algorithmus sich auf hochgradig nichtlineare Systeme bisher nicht anwenden lässt und eine geringe Dateneffizienz aufweist. / In this thesis, the application of reinforcement learning for the control of nonlinear dynamical systems is researched. At first, the relevant principles of stochastic optimal control and machine learning are explained. Afterwards, reinforcement learning is embedded in the context of optimal control. Three methods of deep reinforcement learning are analyzed. A particular algorithm, namely Deep-Deterministic-Policy-Gradient (DDPG), is chosen for further studies on a variety of mechanical systems. Furthermore, the reinforcement learning approach is compared to a model-based trajectory optimization method, called iterative linear-quadratic regulator (iLQR). All control problems can be successfully solved with the trajectory optimization approach, but for new initial conditions, the problem has to be solved again. In contrast, with DDPG a \emph{global} feedback controller is learned, that can drive the controlled system in the desired state. Disadvantageous is the poor data efficiency and the lack of applicability to highly nonlinear systems.