Microelectronic Systems

The ongoing miniaturization as well as advances in packaging technology paves the way to more powerful, intelligent and complex electronic systems, which are used in many applications. Furthermore new ideas to apply modern manufacturing technology enable new sensor concepts. This paper summarizes the work of the Fraunhofer IIS in the field of design of electronic systems. Current challenges of design for manufacturability and reliability as well as design of highly complex systems are discussed and directions are given for exploitation of the advantages in manufacturing technology.

Sub-wavelength gratings and hole arrays in metal films are applicable for polarization and spectral selective sensors, respectively. We demonstrate the fabrication of wire grid polarizers using standard complementary metal–oxide semiconductor (CMOS) processes. Extraordinary optical transmission of hole arrays was achieved by using the dedicated layer of a modified CMOS process. The structures were simulated using the finite-difference time-domain (FDTD) method and fabricated using the work flow of integrated circuits. A high-speed polarization image sensor with a pixel size of 6 nm was designed and demonstrated, and multispectral sensing was implemented using nanostructures with different spectral filter performances on a single chip.

The technology of vertical integration using Through Silicon Vias (TSVs) now is mature for commercial products with a smaller form factor, better performance, less power consumption and lower cost. This paper addresses two challenges faced with the design using vertical integration. First, methods for the characterization of the physical behavior of the new interconnect structures are described. Second, since issues of reliability and thermal management become more important and difficult and the design space is drastically larger, new early stage design tools are needed. A new floorplanning flow is introduced which supports the cost and performance optimized implementation of digital systems in a stack.

Analog to Digital Converters (ADCs), key functions for mixed signal ASICs and SOCs stay in the main focus of our work as enabling technology for many different system applications. New requirements of system as well as of technology aspects call for new concepts for analog to digital conversion like direct sensor signal conversion or digital assisted analog functions. Four examples from current research and development projects are depicted in this article.

Today’s position measurement systems are restricted to systems based on lateral Hall sensors and, hence, suffer from low robustness concerning temperature dependency and disturbing external magnetic fields. With the application of 3-D Hall sensors new principles in signal conditioning can be applied. Position values are then calculated from the complete flux vector and not only from a single component of the magnetic field. Thus it is possible either to perform multi axis position measurement or to use gradient approaches to come to more robust systems. The drawback of these approaches is that the complexity of system design increases. So it is necessary to set up new developmentmethodologies for such multidimensional systems. In this article an approach to design such robust multidimensional position sensors systems is proposed.

The exact determination of displacement and rotation of moving parts is a frequent task in automotive and industrial automation. Systems consisting of permanent magnets and magnetic flux sensors provide the advantage of a precise, robust, non-contact, i.e. wear-free measurement method. Magnetic sensors based on HallinOne® technology enable the design of systems which make use of the entire (3-D) set of magnetic flux density for the determination of up to six relative translational and rotational position coordinates. However, design of corresponding multi-axis magnetic encoder applications poses a quite complex task for the engineer. In this contribution, we explain the design process for multi-axis magnetic position sensing systems and present an inverse pendulum solution as an example.

Modern wireless receivers tend to decrease power consumption. This section shows the benefits of a sub-10µA wake-up receiver circuit used for wireless geofencing and localisation applications with very low maintenance. The proposed wake-up receiver is a wireless receiver which continuously scans the radio channel for certain wake-up messages. Having received and decoded such a message containing additional data, the wake-up receiver triggers different actions in smart objects. The wake-up receiver consumes only 7.5 Microwatts and is suitable for mobile battery-operated or solar-cell driven smart objects. It is shown how the wake-up receiver can be used to implement a wireless in-door geofencing system. This application profits the most from both the very low deterministic current consumption and the short reaction time below 500 ms.

A new low cost antenna for high-precision applications in Global Navigation Satellite Systems (GNSS) is described. The antenna has been designed to operate for all existing satellite navigation frequencies in L-band including GPS, GLONASS, Galileo and Compass (1164–1300MHz and 1559–1610MHz). The right-hand circularly polarized radiating structure consists of a broadband matched square patch with a fourpoint feeding. The antenna exhibits a 10 dB beamwidth of more than 150ı, an antenna gain of more than 3 dBic and an excellent phase center stability. The desired radiation pattern roll-off and antenna gain are obtained using an electrically small radiating element with parasitic elements arranged around it. The dimensions of the developed antenna are a diameter of 146mm and a height of 31 mm.

A cognitive-radio (CR) radio-frequency (RF) receiver front-end architecture is proposed that allows for aggregation of spectral whites spaces. The architecture is able to exploit white spaces that are possibly spread over a wide bandwidth. Key components of the front-end are identified, having considerable impact on the performance of a CR system. Example implementations of those are discussed, including a dual-polarised dual-band antenna, a broadband low-noise amplifier LNA, and a tuneable pre-selection filter.

The modelling language Modelica has been developed by a steadily growing community since 1996. Today, it is a well-established element of the design process in several technical application areas. Fraunhofer IIS/EAS has been involved in the language development from its very beginning. Research on dedicated fields like structural variability and control algorithms usingModelica as well as the development of complete packages (Modelica Standard Library, Complex and Statistics Library) are highlights of these Modelica activities. This article presents a compact overview about the variety of research results which have been achieved by the Dresden Division of Fraunhofer IIS during the last 15 years.

Communication and navigation have always been essential to the success of human societies. The term information society expresses the prominent relevance of information systems in general, and communications and navigation systems in particular for the society of today. Radio communications and radio navigation have a history of more than 100 years. Pioneers like Guglielmo Marconi and Christian Hülsmeyer, the inventor of radar, paved the way to modern information systems together with many others. For more than 25 years Fraunhofer IIS has contributed its share to the rapid development of information systems, concurrently training many young engineers in this seminal field.

This chapter assesses the choice of the physical layer parameters (code rate, modulation, time interleaving) that are best suited for satellite broadcast. The optimum choice depends on the system’s envisaged spectral efficiency or available signal-to-noise ratio, and the target usage environment, like rural or sub-urban reception. The analysis is carried out by means of information theory and evaluation of satellite field measurements.

Energy efficient broadcasting of mobile-TV content through DVB-H and DVB-SH suffers from signaling limitations introduced by the DVB-H and DVB-SH multiplexing scheme resulting in an increased system delay. This paper presents a novel scheduling and multiplexing approach to overcome these limitations. The proposed algorithm allows to trade time slicing efficiency for lower system delay and more flexible data rate assignments possibly resulting in better video quality. The algorithm is evaluated through extensive cross layer simulations for different parameter sets.

The acronym SafeTRIP stands for “Satellite Application for Emergency handling, Traffic alerts, Road safety and Incident Prevention”. Being a European (FP7) project, it brings together 20 partners from seven countries having the common goal of improving road safety, mobility, and environmental protection for all kinds of passenger vehicles. The general objective of the SafeTRIP project is to improve the use of road transport infrastructures and to improve the alert chain (information/prevention/intervention) in case of incidents by offering an integrated system from data collection to service provision. SafeTRIP directly contributes to the achievement of the European Commission’s objectives regarding road transport safety and road mortality reduction. The SafeTRIP system basically is an Intelligent Transport System (ITS) aiming at the combination of digital broadcasting technology with satellite and complementary communication channels to enable novel bi-directional service offerings to the automotive user with pan-European coverage. The satellite will provide seamless coverage for mobile reception whereas the terrestrial components are mainly used to increase the high system capacity in areas with many users or it will increase the quality of service (QoS) in areas with limited satellite coverage. The SafeTRIP project focuses on the test of the core technologies for different applications and usage scenarios. An open architecture is created allowing the development of innovative applications by leveraging a set of enabling services. The demonstrator will support communication via a real satellite, corresponding ground stations or other 3G/4G infrastructure, an on-board unit and end-to-end applications to evaluate the benefits of the architecture and the applications during a trial phase.

The increasing demand for higher data rates, smaller antenna apertures, or less power at the uplink for mobile devices requires air-interface and application specific processing, especially for telecommunication satellites. Only on satellites for dedicated applications or with a short limited lifetime, on-board processing is partly used, but processing on-board improves the system performance or increases the system capacity in several cases. Today’s on-board processing for satellite communication is mostly based on ASIC (Application Specific Integrated Circuit) chips, which have their main drawback in the limited flexibility. In order to demonstrate and validate the flexibility of an FPGA-based on-board processor (OBP) for space applications the Fraunhofer IIS is involved in a so-called in-orbit verification (IOV) payload on the Heinrich-Hertz-Satellite. During the development of an on-board processor for space applications, the main challenge is to ensure a typical life time of 15 years for the hard-, firm- and software under the given environmental conditions. Alternatively, an FPGA platform can be reconfigured for novel communication protocols. In order to investigate new standards for telecommunication satellite systems, the Fraunhofer IIS is developing an OBP platform, based on four FPGAs. The OBP will be embedded into the H2Sat satellite, which will be launched in 2016 and will be located on a geostationary earth orbit (GEO). To the best of our knowledge, this is the first completely reconfigurable platform based on four leading-edge radiation-hardened FPGAs in the geosynchronous orbit for telecommunication satellites. Alexander

In this paper, the implementation of a multi-constellation, multi-frequency automotive GNSS receiver is discussed. The main objective of this paper is threefold. First, to identify, in the context of automotive applications, the most promising GNSS signal combination and analyze its benefits and limitations. Second, to propose a receiver architecture that offers sufficient robustness and flexibility to maintain high-accuracy and high-availability navigation capabilities in challenging automotive signal environments as well as to accommodate the particulars of the legacy, new and modernized signals. Third, to optimize the receiver’s implementation so that it meets the automotive requirements in terms of size, cost and power consumption. To this end, several front-end architectures are compared and some key aspects of the baseband hardware implementation discussed. Additionally, robust acquisition and tracking algorithms that respectively account for the availability of a second frequency and for the introduction of advanced modulations are presented. Finally, some insights are provided regarding optimization of the PVT performance in terms of multipath mitigation and ionospheric corrections.

An approach for pedestrian navigation in indoor environments is presented. It addresses mobile platforms with low processing power and low-cost sensors. Outdoors the horizontal attitude of a device can be easily detected using electronic compasses. Indoors magnetic disturbances lead to unreliable compass outputs. In this paper a novel approach for attitude and position tracking is introduced. Four horizontally arranged directional antennas are used to collect the Wi-Fi signal strengths of transmitters (access points) in range. For attitude estimation an extended Kalman filter is used, and for position tracking Wi-Fi fingerprinting. With this approach the attitude of a mobile device can be estimated and the position can be tracked in indoor environments like e.g. museums. This enables the use of electronic guides that offer additional information by means of augmented reality on exhibits in visual range. Possible accuracies are evaluated in simulations. A test with measurements collected in a museum demonstrates the functionality of the approach.

It is well known, that regular physical activity is an important factor for preserving the health status, the challenge is how to quantify it accurately. Similarly rehabilitation programs rely on physical exercises, where the best results can be achieved through daily training. But who monitors and evaluates exercise execution at home? Micro electro mechanical systems (MEMS) based accelerometers provide a technological solution for inexpensive monitoring systems. The required number of accelerometers within the monitoring system depends on the use case. Quantifying certain activities like walking or cycling can be achieved with only one sensor, while recognizing differences in movements requires more observations and thus a network of accelerometers. We present some typical movement related applications. For these use cases single sensor and multi-sensor systems are compared with respect to their advantages, challenges and limitations. Even signal processing requirements differ from application to application. Two approaches are explained in detail: knowledge-based and model-driven algorithms. While knowledge-based algorithms rely on feature extraction and an inference machine, which infers high level information from these features, model-driven algorithms try to describe relations between collected data and movements.

In this paper a radio-based, real-time tracking system and its application in sports is presented. The system is capable of tracking simultaneously a large number of high dynamic objects in a pre-defined area of interest like a sports stadium. Position and sensor data are captured with high precision and at a high level of detail. Moreover, the position data is automatically analyzed and interpreted from a sports science point of view and presented in a graphical user interface to players, teams and coaches. As an example, a standardized performance test for footballers of the German football association DFB is implemented and presented here. Thus, many disciplines from radio frequency and communication technologies, media engineering, sports medicine and sports science as well as valuable input from sports professionals contributed to this research work. The paper summarizes the manifold results of the research work performed all along the closed processing chain from data capturing, analyzing, condensation and interpretation to its presentation in a 3-D graphical user interface.

Diverse identification and positioning technologies like radio-frequency identification (RFID), real-time locating systems (RTLS), and wireless sensor networks (WSN) are among the enablers for the Internet of Things (IoT). Although the cost-effective integration of these enabling technologies into existing enterprise infrastructures is very important for companies, integration aspects are ignored by researches frequently. Even worse, the increasing diversity of identification and positioning technologies and corresponding applications leads to more expensive integration and application development processes. Up to now these technologies are used rather separately in different applications, but it is only a question of time before these technologies have to be used within the same application in parallel. Furthermore it could be observed, that existing commercial middleware products aiming at integration are focusing on passive RFID only. Hence there is a need to functionally extend existing solutions. This contribution introduces an integration and application platform for diverse enabling technologies for the Internet of Things that has been derived from the requirements of real world applications. An overview on applications of the Internet of Things is given and a selection of applications that have been successfully implemented in research and development projects is examined in detail. The functionalities provided by the platform and the corresponding abstract software architecture are derived from these applications.

Since its foundation over 25 years ago, the Fraunhofer IIS has been involved in the applied research within the field of “Visual Computing”. This field of research incorporates all areas of image acquisition: including different kinds of imaging physics, various types of image sensors; image formation: describing the manipulation of the acquired image data between the sensor and the output and 3-D-image reconstruction; image coding and transportation: lossy and lossless coding, wire-based and wireless transmission; image processing: image-to-image operations such as image enhancement, panoramic image stitching, or video tracking; image analysis: pattern recognition, object detection, classification; and image synthesis: including computer graphics, image augmentation and modeling.

Cost and competition force modern hematology laboratories to further automate their processes. To that respect the examination and analysis of the peripheral blood is of central importance as it is relevant to a large variety of diseases while on the other hand financial reimbursement is low. Over the past eight years, the HemaCAM system has been developed by the Fraunhofer IIS, which supports the assessment of peripheral blood samples and the so-called white blood differential. Since 2010, HemaCAM has been available on the market as a certified medical product, to be more specific as an in vitro diagnostic device. This contribution provides an overview of the key components of the HemaCAM system.

An approach enabling the detection, tracking and fine analysis (e.g. gender and facial expression classification) of faces using a single web camera is described. One focus of the contribution lies in the description of the concept of a framework (the so-called Sophisticated High-speed Object Recognition Engine – SHORE), designed in order to create a flexible environment for varying detection tasks. The functionality and the setup of the framework are described, and a coarse overview about the algorithms used for the classification tasks will be given. Benchmark results have been obtained on both, standard and publicly available face data sets. Even though the framework has been designed for general object recognition tasks, the focus of this contribution lies in the field of face detection and facial analysis. In addition a demonstration application based on the described framework is given to show analysis of still images, movies or video streams.

Understanding plant growth and analyzing plant interaction with the environment is an important aspect in modern agronomy and biological sciences.While measurements are often taken at field scale, current research focuses increasingly on individual plants. As a manual determination of morphological plant parameters is very time-consuming, automatic acquisition methods at high throughput are required. Optical scanning methods provide fast acquisition of surface points. However, as plants represent geometrically complex objects, planning a proper measurement setup and evaluation of the acquired data is a challenging task. This paper addresses solutions for system design and data processing for the sheet-of-light measurementmethod. As an example implementation a 3-D scanning system for individual plants is presented, which is amended by color images for high resolution surface and color measurements of individual plants.

The task was to measure ice cores with a diameter of 10 cm and a length of 1m that were drilled out of Arctic and Antarctic glaciers down to 3000m depth. By means of computed tomography (CT), pieces of 1m length are measured and three-dimensional volume data with high spatial resolution are reconstructed. Complex image processing algorithms are applied to analyze the volume regarding its mean porosity as a function of depth, its total mean porosity and the volumetric distribution of the pores. Since high image quality is required to achieve precise results, a vast amount of data is acquired. The geometry of the samples increases the requirements for the X-ray components. Furthermore measurements have to be made in an environment of -15 °C.

Pistons for combustion motors must meet strongest requirements in terms of material quality and dimensional accuracy to withstand the load inside the combustion chamber. The process integrated inspection of pistons using ultrasonic, eddy current and radiography is therefore state of the art. One drawback of all these methods is the lack of precise defect localisation which results in false rejects. Computerized Tomography (CT) provides a three dimensional characterization of defects and overcomes the above mentioned disadvantage but is established only for sample testing in laboratory. In this contribution we present the realisation of a CT system for process integrated casting inspection by means of combustion motor pistons.We cover the hardware setup as well as the software setup and focus on the reference data comparison which is the key method for process integrated defect detection.

Traditionally, image and video processing algorithms start from an RGB image. However, current image sensors deliver camera raw data that needs additional processing and interpolation for conversion into an RGB representation. While recent research delivers important improvements in image quality for processing and reconstruction of raw images, these algorithms come with a heavy computational complexity. Consequently, they are not suited for mobile solutions such as cameras for media production, mobile phones or surveillance. Offline processing on the other hand offers both higher computational power and better flexibility and is well suited for executing those algorithms. The workflow for utilizing this enhanced quality thus requires a shift from camera centric imaging to new off-camera processing strategies. This requires a novel infrastructure for transportation and interchange and enables the possibility for development of even more sophisticated algorithms for processing of camera raw data. This contribution highlights the challenges and possibilities arising from the above mentioned paradigm shift. We discuss algorithms that should stay within the camera and algorithms that benefit from offloading. Our research contributes to the increase in image quality of workflows for future video applications.

As an essential and renowned business field of Fraunhofer IIS, the audio and multimedia activities offer advanced technologies that enable a broad range of multimedia enabled applications. While this may happen in many different contexts, such as Internet transmission, consumer devices or mobile multimedia services of next-generation communication, it is always the quality experienced by the human listener/viewer that constitutes the essential criterion of merit which several generations of technology are striving to fulfill to an ever-increasing extent. Naturally, since the intended final receiver is the human auditory/visual system in this case, the quality experienced is subjective in nature and requires a profound understanding in psychometrics, psychoacoustics and psychovisual disciplines for its successful measurement and optimization.

Unlike loudspeakers, headphones are not designed to have a flat frequency response. Instead they should compensate for the spectral coloration caused by the acoustic transfer path when sound is travelling from a point in space to the ear. For correct headphone reproduction it is essential to control the sound pressure at the listener’s ears. In the literature, there is no consensus about the optimal transfer function and equalization of headphones. In this work, several equalization strategies were tested on different commercially available headphones. Headphones were measured on an artificial head and equalization filters were designed in the frequency domain consisting of two parts: The first part of the filter is specific for each headphone and linearizes the frequency response of the headphone at the point of measurement, i.e. at the beginning of the blocked ear-channel. The second part of the filter is generic for all headphones and allows testing of different target responses on different headphones. Target responses from literature and the ISO 11904-1 diffusefield equalization were tested. Further, target responses were designed by expertlisteners by directly comparing headphone and loudspeakers playback. A listening test showed a general preference for the equalization curves adjusted by experts and the diffuse-field equalization. The experiment also verified that the equalization process enhances the perceptual quality irrespective of the headphone model.

The recording and reproduction of spatial audio gains increasing attention since multichannel sound systems have been established in modern audio applications. Directional Audio Coding (DirAC) represents an efficient description of spatial sound in terms of a few audio downmix signals and parametric side information in frequency sub-bands, namely the direction-of-arrival and diffuseness of the sound. Traditionally, the directional parameters are derived based on the active sound intensity vector, which can be efficiently estimated via two or threedimensional microphone grids. However, due to form factor constraints, linear microphone arrays are often more suitable in practice and thus, alternative parameter estimation techniques are required. In this contribution, a framework for parametric spatial audio processing based on linear microphone arrays is presented allowing for an efficient recording, enhancement, and reproduction of spatial sound. Different application examples are discussed including parametric directional filtering and de-reverberation.

The European research project “Together anywhere, together anytime” (TA2) explores how technology can make communication and engagement easier among groups of people separated in space and time, e.g. families living in different cities. It combines high quality communication with shared group activities using distributed multimedia applications and broadband technology in the living room. This paper provides an overview of the TA2-project with a focus on highdefinition audio communication, which has the goal of providing an experience “as if speaking to someone in the same room”. The most important aspects for making audio communication a natural experience are high audio bandwidth, robust transmission with low delay, hands-free operation, and spatial audio (surround sound) reproduction. These requirements are addressed by the Audio Communication Engine (ACE), which includes audio coding, acoustic processing and IP transport in a single software module. Based on concept demonstrators developed in the TA2- project, results from user experience tests are presented which indicate that more than 90% of users rate the audio quality as natural when using high-definition audio, compared to less than 10% when standard audio quality is used.

Recently MPEG finished the standardization of a Low Delay MPEG Surround tool that is tailored for enhancing the widely adopted AAC-ELD low delay codec for high-quality audio communication into AAC-ELD v2. In combination with the Low Delay MPEG Surround tool, the coding efficiency for stereo content outperforms competing low delay audio codecs. This paper describes the technical challenges and solutions for designing a low delay codec that delivers a performance which is comparable to that of existing state of the art compression schemes. It provides a comparison to competing proprietary and ITU-T codecs, as well as a guideline for how to select the best possible points of operation. Applications facilitated by AAC-ELD v2 in the area of broadcasting and mobile video conferencing are discussed.

Speech and audio coding schemes originate from different worlds. Speech coding schemes typically assume a source model i.e. the human vocal tract. General audio coding schemes primarily rely on a sinkmodel i.e. the human auditory system. While speech coding schemes work well for the signal class they were designed for at very low rates, they are known to fail for general audio signals even at higher rates. In contrast, general audio coders work well for any content at higher rates, but typically have limited performance especially for speech signals at very low rates. Recently the ISO/MPEG group started a standardization activity to develop a new Unified Speech and Audio Coding scheme. A state of the art AAC based general audio coder, featuring transform coding, parametric bandwidth extension and parametric stereo coding,was extended by source model coding tools. All codec modules were further improved and revised for enhanced performance in particular at very low bitrates. The new unified coding scheme outperforms dedicated speech and general audio coding schemes and bridges the gap between both worlds. This paper describes the new codec in detail and shows how the goal of consistent high quality for all signal types is reached.

Parametric spatial coding of multichannel content has acquired widespread use in today’s audio codecs to enable efficient operation in the medium bitrate range. However, applause signals are still challenging to code with good perceptual quality using these techniques. To obtain a sufficiently low side information bitrate, the spatial parameter update rate is usually chosen in the range of several tens of milliseconds which is too low to restore a convincing spatial listener envelopment of applause-like signals. Moreover, state of the art decorrelators that are mandatory in parametric multichannel decoders inevitably deteriorate the signal quality of transient sound events, like handclaps, by dispersion. Therefore, we propose a novel decorrelation and parametrization technique for efficient coding of applause sounds: transient events are separated from the core decoder output and a dedicated decorrelator algorithm distributes the transients in the spatial image according to parametric guiding information transmitted in the bitstream. Listening tests show a substantial improvement in subjective quality compared to standard methods.