Algorithm design methods for radio communications systems are poised to undergo a massive disruption over the next several years. Today, such algorithms are typically designed manually using compact analytic problem models. However, they are shifting increasingly to machine learning based methods using approximate models with high degrees of freedom, jointly optimized over multiple subsystems, and using real-world data to drive design which may have no simple compact probabilistic analytic form. Over the past five years, this change has already begun occurring at a rapid pace in several fields. Computer vision tasks led deep learning, demonstrating that low level features and entire end-to-end systems could be learned directly from complex imagery datasets, when a powerful collection of optimization methods, regularization methods, architecture strategies, and efficient implementations were used to train large models with high degrees of freedom. Within this work, we demonstrate that this same class of end-to-end deep neural network based learning can be adapted effectively for physical layer radio systems in order to optimize for sensing, estimation, and waveform synthesis systems to achieve state of the art levels of performance in numerous applications. First, we discuss the background and fundamental tools used, then discuss effective strategies and approaches to model design and optimization. Finally, we explore a series of applications across estimation, sensing, and waveform synthesis where we apply this approach to reformulate classical problems and illustrate the value and impact this approach can have on several key radio algorithm design problems. / Ph. D. / Radio communications and sensing systems are used pervasively in the modern world every day life to connect phones, computers, smart devices, industrial devices, internet services, space systems, emergency and military users, radar systems, interference monitoring systems, defense electronic systems, and others. Optimizing these systems to function together reliably and efficently in an ever more complex world is becoming increasingly hard and impractical. Our work introduces a new and radically different method for the design of radio systems by casting them in a new way as artificial intelligence problems relying on the field of machine learning called deep learning to find and optimize their design. We detail and demonstrate the first such deep learning based communciations and sensing systems operating on raw radio signals and quantify their performance when compared to existing methods, showing them to be competitive with and in some cases significantly better performing than state of the art systems today. These ideas, and the evidence of their viability, are central to the emerging field of machine learning communications systems, and will help to make tomorrow’s wireless systems faster, cheaper, more reliable, more adaptive, more efficient, and lower power than currently possible. In a world of ever increasing complexity and connectedness, this new approach to wireless system design from data using machine learning offers a powerful new strategy to improve systems by directly leveraging the complexity in real world data and experience to find efficiencies where current day approaches and insufficient simplified models and design tools can not.

Despite the fact ultra-wideband (UWB) technology has been around for over 30 years, there is a newfound excitement about its potential for communications. With the advantageous qualities of multipath immunity and low power spectral density, researchers are examining fundamental questions about UWB communication systems. In this work, we examine UWB communication systems paying particular attention to transmitter and receiver design. This thesis is specifically focused on a software radio transceiver design for impulse-based UWB with the ability to transmit a raw data rate of 100 Mbps yet encompasses the adaptability of a reconfigurable digital receiver. A 500 ps wide Gaussian pulse is generated at the transmitter utilizing the fast-switching characteristics of a step recovery diode. Pulse modulation is accomplished via several stages of RF switches, filters, and amplifiers on a fully designed printed circuit board specifically manufactured for this project. Critical hardware components at the receiver consist of a bank of ADCs performing parallel sampling and an FPGA employed for data processing. Using a software radio design, various modulation schemes and digital receiver topologies are accommodated along with a vast number of algorithms for acquisition, synchronization, and data demodulation methods. Verification for the design is accomplished through transmitter hardware testing and receiver design simulation. The latter includes bit error rate testing for a variety of modulation schemes and wireless channels using a pilot-based matched filter estimation technique. Ultimately, the transceiver design demonstrates the advantages and challenges of UWB technology while boasting high data rate communication capability and providing the flexibility of a research testbed. / Master of Science

With the explosive development of wireless communication systems the specifications of the supporting hardware platforms have become more and more demanding. According to the long term goals of the industry, future communications systems should integrate a wide variety of standards. This leads to the idea of software defined radio, implemented on fully reconfigurable hardware.Among other reconfigurable hardware blocks, suitable for the software radio concept, an outstanding importance belongs to the reconfigurable filters that are responsible for the selectivity of the system. The problematic of filtering is strictly connected to the architecture chosen for a multi-mode receiver realization. According to the chosen architecture, the filters can exhibit low pass or band pass frequency responses.The idea of reconfigurable frequency parameters has been introduced since the beginning of modern filtering applications due to the required precision of the frequency response. However, the reconfiguration of the parameters was usually done in a limited range around ideal values. The purpose of the presented research is to transform the classical filter structures with simple self-correction into fully reconfigurable filters over a wide range of frequencies. The ideal variation of the frequency parameters is continuous and consequently difficult to implement in real circuits. Therefore, it is usually sufficient to use a discrete programming template with reasonably small steps.There are several methods to implement variable frequency parameters. The most often used programming templates employ resistor and capacitor arrays, switched according to a given code. The low pass filter implementation proposed in this work uses a special switching template, optimized for a quasi-linear frequency variation over logarithmic axes. The template also includes the possibility to compensate errors caused by component tolerances and temperature. Another important topic concerns the implementation of programmable band pass filters, suitable for IF sampling receivers. The discussion is centered on the feasibility and the flexibility of different band pass filter architectures. Due to the high frequency requirements, the emphasis lays on filters that employ transconductance amplifiers and capacitors. / Die rasch fortschreitende Entwicklung drahtloser Kommunikationssysteme führt zu immer anspruchsvolleren Spezifikationen der diese Systeme unterstützenden Hardwareplattformen. Zukünftige Kommunikationssysteme sollen übereinstimmend mit den längerfristigen Zielen der Industrie verschiedene Standards integrieren. Dies führt zu der Idee von vollständig rekonfigurierbarer Hardware, welche mittels Software gesteuert wird.Inmitten anderer rekonfigurierbarer Hardwareblöcke, die für das Software Radio Konzept geeignet sind, besitzen die steuerbaren Filter, welche wesentlichen Einfluss auf die Selektivität des Systems haben, eine enorme Bedeutung. Die Filterproblematik ist eng mit der gewählten Architektur der standardübergreifenden Empfängerrealisierung verknüpft. Die Filter können entsprechend der ausgesuchten Architektur Tiefpass- oder Bandpasscharakter annehmen.Die Idee rekonfigurierbarer Frequenzparameter wurde bereits mit Beginn moderner Filteranwendungen auf Grund geforderter Frequenzganggenauigkeit umgesetzt. Jedoch wurde die Parameterrekonfiguration üblicherweise nur in einem begrenzten Bereich um die Idealwerte herum vorgenommen. Das Ziel der vorgestellten Forschungsarbeit ist es, diese klassischen Filterstrukturen mit einfacher Selbstkorrektur in über große Frequenzbereiche voll rekonfigurierbare Filter zu transformieren. Idealerweise werden die Frequenzparameter kontinuierlich variiert weswegen sich die Implementierung in reellen Schaltkreisen als schwierig erweist. Deshalb ist es üblicherweise ausreichend, ein diskretes Steuerschema mit kleinen Schrittweiten zu verwenden.Es gibt verschiedene Methoden, variable Frequenzparameter zu implementieren. Die meisten Schemata verwenden Widerstands- und Kondensatorfelder, die entsprechend eines Kodes geschaltet werden. Die in dieser Arbeit vorgestellte Implementierung eines Tiefpassfilters nutzt ein spezielles Umschaltschema, welches für die quasi-lineare Frequenzvariation bei Darstellung über logarithmischen Axen optimiert wurde. Es beinhaltet weiterhin die Möglichkeit, Fehler zu kompensieren, die durch Bauelementtoleranzen und Temperaturschwankungen hervorgerufen werden.Ein weiteres interessantes Thema betrifft die Implementierung steuerbarer Bandpassfilter, die für Empfänger mit Zwischenfrequenzabtastung geeignet sind. Die Betrachtung beschränkt sich hierbei auf die Durchführbarkeit und Flexibilität verschiedener Bandpassfilterarchitekturen. Auf Grund hoher Frequenzanforderungen liegt der Schwerpunkt auf Filtern, die auf Transkonduktanzverstärkern und Kondensatoren basieren.

With the explosive development of wireless communication systems the specifications of the supporting hardware platforms have become more and more demanding. According to the long term goals of the industry, future communications systems should integrate a wide variety of standards. This leads to the idea of software defined radio, implemented on fully reconfigurable hardware.Among other reconfigurable hardware blocks, suitable for the software radio concept, an outstanding importance belongs to the reconfigurable filters that are responsible for the selectivity of the system. The problematic of filtering is strictly connected to the architecture chosen for a multi-mode receiver realization. According to the chosen architecture, the filters can exhibit low pass or band pass frequency responses.The idea of reconfigurable frequency parameters has been introduced since the beginning of modern filtering applications due to the required precision of the frequency response. However, the reconfiguration of the parameters was usually done in a limited range around ideal values. The purpose of the presented research is to transform the classical filter structures with simple self-correction into fully reconfigurable filters over a wide range of frequencies. The ideal variation of the frequency parameters is continuous and consequently difficult to implement in real circuits. Therefore, it is usually sufficient to use a discrete programming template with reasonably small steps.There are several methods to implement variable frequency parameters. The most often used programming templates employ resistor and capacitor arrays, switched according to a given code. The low pass filter implementation proposed in this work uses a special switching template, optimized for a quasi-linear frequency variation over logarithmic axes. The template also includes the possibility to compensate errors caused by component tolerances and temperature. Another important topic concerns the implementation of programmable band pass filters, suitable for IF sampling receivers. The discussion is centered on the feasibility and the flexibility of different band pass filter architectures. Due to the high frequency requirements, the emphasis lays on filters that employ transconductance amplifiers and capacitors. / Die rasch fortschreitende Entwicklung drahtloser Kommunikationssysteme führt zu immer anspruchsvolleren Spezifikationen der diese Systeme unterstützenden Hardwareplattformen. Zukünftige Kommunikationssysteme sollen übereinstimmend mit den längerfristigen Zielen der Industrie verschiedene Standards integrieren. Dies führt zu der Idee von vollständig rekonfigurierbarer Hardware, welche mittels Software gesteuert wird.Inmitten anderer rekonfigurierbarer Hardwareblöcke, die für das Software Radio Konzept geeignet sind, besitzen die steuerbaren Filter, welche wesentlichen Einfluss auf die Selektivität des Systems haben, eine enorme Bedeutung. Die Filterproblematik ist eng mit der gewählten Architektur der standardübergreifenden Empfängerrealisierung verknüpft. Die Filter können entsprechend der ausgesuchten Architektur Tiefpass- oder Bandpasscharakter annehmen.Die Idee rekonfigurierbarer Frequenzparameter wurde bereits mit Beginn moderner Filteranwendungen auf Grund geforderter Frequenzganggenauigkeit umgesetzt. Jedoch wurde die Parameterrekonfiguration üblicherweise nur in einem begrenzten Bereich um die Idealwerte herum vorgenommen. Das Ziel der vorgestellten Forschungsarbeit ist es, diese klassischen Filterstrukturen mit einfacher Selbstkorrektur in über große Frequenzbereiche voll rekonfigurierbare Filter zu transformieren. Idealerweise werden die Frequenzparameter kontinuierlich variiert weswegen sich die Implementierung in reellen Schaltkreisen als schwierig erweist. Deshalb ist es üblicherweise ausreichend, ein diskretes Steuerschema mit kleinen Schrittweiten zu verwenden.Es gibt verschiedene Methoden, variable Frequenzparameter zu implementieren. Die meisten Schemata verwenden Widerstands- und Kondensatorfelder, die entsprechend eines Kodes geschaltet werden. Die in dieser Arbeit vorgestellte Implementierung eines Tiefpassfilters nutzt ein spezielles Umschaltschema, welches für die quasi-lineare Frequenzvariation bei Darstellung über logarithmischen Axen optimiert wurde. Es beinhaltet weiterhin die Möglichkeit, Fehler zu kompensieren, die durch Bauelementtoleranzen und Temperaturschwankungen hervorgerufen werden.Ein weiteres interessantes Thema betrifft die Implementierung steuerbarer Bandpassfilter, die für Empfänger mit Zwischenfrequenzabtastung geeignet sind. Die Betrachtung beschränkt sich hierbei auf die Durchführbarkeit und Flexibilität verschiedener Bandpassfilterarchitekturen. Auf Grund hoher Frequenzanforderungen liegt der Schwerpunkt auf Filtern, die auf Transkonduktanzverstärkern und Kondensatoren basieren.

International Telemetering Conference Proceedings / October 21, 2002 / Town & Country Hotel and Conference Center, San Diego, California / Waveform Description Language (WDL) was invented to ease the process of porting legacy and/ or new radio waveforms to Programmable / Software Radios. WDL has two primary requirements; 1st it is to provide a rigorous executable behavioural description of waveform signal structures that is unambiguous and yet independent of any particular end item software radio architecture. The 2nd requirement is that the behavioural specification provides a path to automatic code generation for GP’s, DSP’s, and FPFG’s and that the Generated code be tested against the behavioural model.

Ionospheric scintillation is a phenomenon caused by varying concentrations of charged particles in the upper atmosphere that induces deep fades and rapid phase rotations in satellite signals, including GPS. During periods of scintillation, carrier tracking loops often lose lock on the signal because the rapid phase rotations generate cycle slips in the PLL. One solution to mitigating this problem is by employing decision-directed carrier recovery algorithms that achieve data wipe-off using differential bit detection techniques. Other techniques involve PLLs with variable bandwidth and variable integration times. Since nearly 60% of the GPS signal repeats between frames, this thesis explores PLLs utilizing variable integration times and decision-directed algorithms that exploit the repeating data as a training sequence to aid in phase error estimation. Experiments conducted using a GPS signal generator, software radio, and MATLAB scintillation testbed compare the bit error rate of each of the receiver models. Training-based methods utilizing variable integration times show significant reductions in the likelihood of total loss of lock.

Zugl.: Dortmund, Techn. Univ., Diss., 2008

L'utilisation des ondes électromagnétiques, en tant que moyen de transmission d'informations entre les terminaux mobiles, a rapidement conduit à une congestion du spectre fréquentiel. Pour améliorer le flux de données entre utilisateurs, les autorités réglementent l'accès aux bandes de fréquences en imposant des normes strictes aux télécommunications mobiles. Ainsi, et afin d'accroître leurs capacités d'échange de données, les prochaines générations de terminaux mobiles doivent utiliser dynamiquement les ressources spectrales. Ces contraintes affectent la conception des émetteurs-récepteurs mobiles. Ils doivent transmettre des débits de données élevés en utilisant plusieurs porteuses ayant différentes modulations, tout en consommant le moins d'énergie possible, afin d'économiser l’autonomie de leurs batteries. L'émetteur Walsh tente de répondre à ces défis / The use of electromagnetic waves, as a medium for transmitting information between the mobile terminals, has quickly lead to a congestion of the frequency spectrum. To improve the data traffic flow between users, the authorities regulate the access to the frequency bands by imposing stringent standards to the mobile telecommunication. Thus, and in order to increase their data exchange capabilities, the next generations of mobile terminals have to dynamically use the spectral resources. These constraints affect the design of mobile transceivers. They must transmit high data rates by using multiple carriers and various modulation schemes, while consuming as less energy as possible, in order to save their battery life. The Walsh transmitter tries to answer these challenges.

This thesis is part of the project undertaken to develop a diffusive acoustic confocal imaging system (DACI) that aims to differentiate between healthy and the diseased tissues in the prostate. Speed of sound is chosen as the tool to quantify the alterations in the tissues’ mechanical properties at different pathological states. The current work presents a scanning configuration that features three components: an acoustic emitter, a focusing mirror and a point receiver. The focusing mirror brings the collimated acoustic beam from the emitter into a focused probe position, which needs to be located within the bladder or at the near surface of the prostate. This position is introduced as the virtual source, where the acoustic intensity diffusively scatters into all directions and propagates through the specimen. The system design was simulated using ZEMAX and COMSOL to validate the concept of the virtual source. Lesions in a phantom prostate were found in the simulated amplitude and phase images. The speed of sound variation was estimated from the 1D unwrapped phase distribution indicating where the phase discontinuities existed. The measurements were conducted in a water aquarium using the tissue-mimicking prostate phantom. Two-dimensional projected images of the amplitude and the phase distributions of the investigating acoustic beam were measured. A USRP device was set up as the signal generation and acquisition device for the experiment. Two different signal extractions methods were developed to extract the amplitude and the phase information. The experimental results were found to generally agree with the simulation results. The proof-of-concept design was successful in measuring both the phase and the amplitude information of the acoustic signal passing through the prostate phantom. In future, the 2D/3D speed of sound variation needs to be estimated by an appropriate image reconstruction method. / Graduate / 2018-12-06

Les systèmes de navigation par satellites permettent les applications de positionnement en extérieur, dont la navigation routière. Dans les environnements contraints, comme l'intérieur des bâtiments où ces signaux satellitaires sont dégradés, la continuité du service de positionnement est nécessaire. Les applications adaptées aux citoyens modernes avec leurs appareils nomades posent des contraintes fortes de mobilité, de coûts et de limitations des infrastructures existantes. Les larges possibilités de déplacements dans des environnements hétérogènes accroissent les difficultés. Un état de l'art alimenté par une décennie de travaux académiques et industriels présente un ensemble de technologies qui visent disponibilité et performance. L'accent porte ensuite sur les systèmes inertiels pédestres à bas coût, avec une première contribution permettant d'abandonner la détection de pas au profit d'une mobilité facilitée, mais reste limitée par la connaissance de la distance parcourue pendant une phase de calibration. Cette approche nouvelle est confrontée à celle classique au pied, puis éprouvée pour différents capteurs et piétons au travers d'expérimentations répétées en conditions réalistes. Une seconde contribution décline une constellation radio locale pour estimer la distance avec une infrastructure allégée à deux émetteurs. Elle s’inspire d’une conception satellitaire sur radio programmable pour faciliter sa compatibilité avec l’existant et explorer ses performances. Une surveillance du rapport signal à bruit inter-canal améliore la précision du positionnement. Le couplage de ces systèmes asynchrones et distribués est évalué en intérieur sur une plateforme automatisée / The global navigation satellite systems allow outdoor positioning applications, including car navigation. In challenging environments, such as the buildings where satellite signals are mitigated, georeferenced points of interest or navigation applications require a continuity of the positioning service. The applications adapted to modern citizens and their mobile devices raise strong constraints on mobility, costs and limitations of the existing infrastructure. The wide variety of displacements in heterogeneous environments increases the challenge. A state of the art fed by a decade of academic and industrial works presents a set of technologies that target availability and performance. The emphasis follows on the low cost pedestrian inertial systems, with a first contribution allowing to give up the step detection for the benefit of an easier mobility, but remains limited to the knowledge of the distance traveled during a calibration phase. This new approach is compared with the classical foot-mounted approach, and then benchmarked with several sensors and pedestrians through repeated experiments in real conditions. A second contribution operates a local radio constellation to estimate the distance with a minimal infrastructure with two emitters. The signals and the algorithm are based on a reproduction of satellite systems to ease the compatibility but are implemented on a programmable radio to explore the performances. A monitoring of the difference of carrier to noise ratio between the radio channels improves the distance estimation. The hybridization of these distributed, asynchronous and multi-rates inertial and radio systems is evaluated indoor on a motorized platform