Le sujet de cette thèse est la conception et la réalisation d'un modem radio COFDM et pouvant être utilisé dans le cadre des applications de la société ADVANTEN.<br />Le premier chapitre est consacré à l'étude des canaux de propagation et aux caractéristiques de la modulation OFDM.<br />Dans le second chapitre, des techniques de synchronisation du récepteur et d'estimation de canal sont présentées. La sélection et le dimensionnement d'une solution adaptée au besoin spécifique sont effectués dans ce chapitre.<br />Dans la suite de l'étude, des algorithmes de codage de canal sont comparés. Nous montrons comment ils permettent une amélioration notable de la qualité de la liaison radio. Des résultats de simulation convaincants sont présentés.<br />Le problème du rapport puissance instantanée à puissance moyenne des signaux OFDM est exposé. Une solution innovante pour réduire ce rapport est proposée.<br />Enfin, les résultats de la phase de réalisation du modem sont donnés dans le chapitre cinq.

Solutions to the problem of large peak to average power ratio (PAPR) in orthogonal frequency division multiplexing (OFDM) systems are proposed. Although the design of PAPR reduction codewords has been extensively studied and the existence of asymptotically good codes with low PAPR has been proved, still no reduced PAPR capacity achieving code has been constructed. This is the topic of the current thesis.This goal is achieved by implementing a time-frequency turbo block coded OFDM. In this scheme, we design the frequency domain component code to have a PAPR bounded by a small number. The time domain component code is designed to obtain good performance while the decoding algorithm has reasonable complexity. Through comparative numerical evaluation we show that our method achieves considerable improvement in terms of PAPR with slight performance degradation compared to capacity achieving codes with similar block lengths. For the frequency domain component code, we used the realization of Golay sequences as cosets of the fi rst order Reed-Muller code and the modi cation of dual BCH code. A simple MAP decoding algorithm for the modi ed dual BCH code is also provided. Finally, we provide a flexible and practical scheme based on probabilistic approach to a PAPR problem. This approach decreases the PAPR without any signi cant performance loss and without any adverse impact or required change to the system. / Engineering and Applied Sciences

An explosive growth in demand for broadband mobile wireless services is currently being fuelled by cellular telephone users who, encouraged by service providers, are no longer content with voice transmission only but are demanding real-time video services, including multi-user, interactive games and 'movie' programmes. As these applications develop, expectations mount in other mobile user markets, especially the public safety arena, for comparable user features but with greater emphasis on reliability and robustness of the equipment and supporting network in adverse propagation conditions, remote locations and emergencies. These applications all have in common the requirements for efficient use of wireless bandwidth and of battery power, as well as seamless operation when moving, sometimes at high vehicle speeds, from one type of environment to another in a multi-user scenario. Orthogonal frequency-division multiplexed (OFDM) signals have been found to compare favourably with other modulation systems in these applications, the multi-carrier format being more tolerant of delay spread. It has been used in both code-division (MC-CDMA) and frequency-division (OFDMA) multi-user schemes, the latter having the advantage of maintaining orthogonality among users in fading-signal environments, with consequent simplification of signal processing. The major drawback of OFDM has been the high peak-to-average power ratio (PAPR) that is characteristic of signals with multiple sub-carriers. A result of this is that the transmitter requires a linear power amplifier (PA) that generally has to be 'backed off' to accommodate the high PAPR. Additional back-off is required to achieve linearity, as well as sometimes-complex linearisation circuitry. The power usage and cost of such a transmitter is more acceptable in a base station, tending to limit the application of OFDM to downlinks. The potential application to hand-portable terminals has severe constraints of size, cost and battery life, exacerbated by the use of video-capable LCD displays, increasing motivation for the use of MIMO (multi-antenna) technology and the development of mobile ad-hoc networks, the latter being particularly applicable in the public safety arena. Previous efforts to ameliorate the PAPR problem have been principally directed at two areas, the reduction of signal PAPR, by block coding, clipping or other techniques, and methods of achieving PA linearisation with improved power efficiency. The first object of the present research was to establish, as far as practicable, the current state of the art in these areas, to set a performance baseline. The next step was to develop an improved transmitter modulation scheme that would not only be able to take advantage of any existing peak reduction methods but would transmit a signal that would be compatible with existing OFDM receivers. A novel modulation technique is now presented, termed Quadrature Phase-Time Modulation (QPTM), that has been found to meet the requirements for linearity, simplicity and low cost, whilst being able to take advantage of constant-envelope PA technology, with its attendant power efficiency. After final amplification, the signal is restored by a passive narrow-band filter to standard OFDM form, having both phase and amplitude modulation. The QPTM system of modulation relies on a dual baseband pulse-width modulation process, performed at a substantially-higher rate than the upper baseband frequency, followed by direct quadrature modulation of a carrier signal. The work undertaken has been in the nature of a feasibility study, commencing with the theoretical basis of the technique, from which a behavioural system model was designed and simulated. After the system was simulated successfully, in several forms, a model was designed for realisation with available high-frequency integrated circuits. From this design, prototypes were constructed and tested. The prototype circuit boards also included an experimental UHF Class-D PA circuit, excluding the output filter, to facilitate ongoing development of the PA and filter subsystem as a separate project. This type of PA was seen as a potential complement to the QPTM modulator, although the technology was at an early stage of development. The prototype PA has a novel push-pull arrangement of GaAs FETs that employs a broadside-coupled tapered-stripline balun instead of the usual transformer. Preliminary measurements were made on the PA using both a spectrum analyser and a newly-available 8GHz-bandwidth digital oscilloscope to confirm basic operating characteristics. The performance of the QPTM technique at frequencies needed for broadband operation is dependent on its practical implementation, which has therefore been a major focus. The inherent difficulties in realising a highly-linear 40MHz triangle-wave reference generator, with a precise ultra-high-speed comparator and modulator system, have been overcome with the chosen design techniques and attention to several critical aspects. The result has been the successful demonstration of QPTM as an efficient PA modulation technique that is equally applicable to either narrow-band, high-capacity UHF or broadband OFDM microwave systems.

In this thesis, reduction of the Peak-to-Average Power Ratio (PAPR) of Orthogonal Frequency Division Multiplexing (OFDM) is studied. A new PAPR reduction method is proposed that is based on block coding the input data and modifying the codeword until the PAPR is reduced below a certain threshold. The method makes use of the error correction capability of the block code employed. The performance of the algorithm has been investigated through theoretical models and computer simulations. For performance evaluation, a new gain parameter is defined. The gain parameter considers the SNR loss caused by modification of the codeword together with the PAPR reduction achieved. The gain parameter is used to compare a plain OFDM system with the system employing the PAPR reduction algorithm. The algorithm performance is examined through computer simulations and it is found that power reductions around 2-3 dB are obtained especially for low to moderate number of channels and relatively strong codes.

Constantly increasing demand for wireless communications in various applications has always led to new ways of modulating the radio frequency (RF) carrier signal by advancing waveform structure throughout generations. Although communication data rates are limited by the theoretical capacity, specific signaling designs for the signal that experiences natural and artificial effects in the transmission medium such as multipath fading channel, hardware impairments and multiuser environment promised better solutions in providing improved wireless access to various type of users and networks. Besides communication capacity, broadcasting nature of radio signals poses the information security as another main concern in wireless communications. In this dissertation, new advanced methods for improving signal statistics in multiple domains are studied. Instead of focusing on a single aspect, the waveform design approaches studied in this dissertation tackle with improving the orthogonal frequency-division multiplexing (OFDM)-based signaling in multiple perspectives such as out-of-band (OOB) emission reduction, peak-to-average-power ratio (PAPR) reduction, and secure transmission with minimum or no eect at the receiver side. Various concepts are coherently exploited while achieving aforementioned goals with minimal cost such as unexplored spaces in the signal space like (CP), guard band, multipath fading; multivariate nature of the multicarrier signals; time spreading and location uniqueness of the wireless channels. The proposed techniques are analyzed theoretically and performance results are presented including related previous works in the literature. It is worth noting that the methods presented in the dissertation can be easily applicable to conventional OFDM systems thanks to having no or minimal change in the receiver structure.

L’OFDM est une technique de modulation multiporteuses largement utilisée dans des applications de communications filaires et sans-fils comme le DVB-T/T2, le Wifi, et la 4G, grâce à sa robustesse contre les canaux sélectifs en fréquence en comparaison avec la modulation monoporteuse. Cependant, le signal OFDM souffre de grandes variations d’amplitude. Les fluctuations de l’enveloppe du signal OFDM génèrent des distorsions non-linéaires quand on introduit le signal dans un équipement non-linéaire comme l’amplificateur de puissance. Réduire les variations du signal améliore le rendement de l’amplificateur, réduit la consommation énergétique et diminue les émissions de CO2 des transmissions numériques.Le PAPR (rapport de la puissance crête sur la puissance moyenne) est une variable aléatoire qui a été introduite pour mesurer les variations du signal. Il existe plusieurs systèmes multiporteuses basés sur différentes bases de modulation et filtres de mise en forme. Nous prouvons d’abord dans ces travaux que le PAPR dépend de cette structure de modulation. Ensuite, nous étudions le comportement du PAPR vis-à-vis des formes d’ondes utilisées dans la modulation. Le problème de réduction du PAPR est ainsi formulé en un problème d’optimisation. Par ailleurs, une condition nécessaire pour construire des formes d’ondes avec un meilleur PAPR que l’OFDM est développée. Cette condition est notamment satisfaite par des bases en ondelettes. Enfin, une nouvelle forme d’onde en paquets d’ondelettes adaptative est proposée, permettant des gains significatifs en PAPR, tout en maintenant les avantages des modulations multiporteuses. / OFDM is a multicarrier modulation system widely used in wireline and wireless applications such as DVB-T/T2, Wifi, and 4G, due to its resilience against frequency selective channels compared with the single carrier modulation systems. However, the OFDM signal suffers from large amplitude variations. The fluctuations of the OFDM envelope generate non-linear distortions when we introduce the signal into a non-linear device like the power amplifier. Reducing the variations of the signal improves the power amplifier efficiency, reduces the energy consumption and decreases CO2 emissions.The peak-to-average power ratio (PAPR) has been introduced as a random variable that measures the power variations of the signal. There exist several multicarrier modulation systems based on different modulation basis and shaping filters. We first prove in this work that the PAPR depends on this modulation structure. Moreover, the behaviour of the PAPR regarding to the modulation waveforms is analysed and the PAPR reduction problem is formulated as an optimization problem. Furthermore, a necessary condition for designing waveforms with better PAPR than OFDM is developed. This necessary condition is particularly satisfied by wavelet basis. Finally, a new adaptive wavelet packet waveform is proposed, allowing significant gain in terms of PAPR, while keeping the advantages of multicarrier modulations.

Cíl disertační práce leží v oblasti modelování a vyhodnocení bezdrátových komunikačních systémů s dvojrozměrným rozprostíráním signálu a jejich klíčových parametrů v závislosti na vybraných vlastnostech moderního bezdrátového komunikačního řetězce. Výzkumné metody použité v této práci spočívají především ve vývoji softwarového simulátoru pro prostředí Matlab, s jehož pomocí, a s využitím statistického přístupu, jsou navržené algoritmy ověřeny. Dále je použit simulátor fyzické vrstvy dle 3rd Generation Partnership Project Long Term Evolution (3GPP LTE), vyvinutý na Technické univerzitě ve Vídni. Tento představuje ideální platformu pro implementaci metody dvojrozměrného (2D) rozprostírání a její vyhodnocení s přihlédnutím k současným bezdrátovým komunikačním systémům. Zjištění prezentovaná v této práci představují především ověření účinnosti systému nazvaného jako Variable Spreading Factor - Orthogonal Code Frequency Division Multiplex (VSF-OFCDM), který využívá principu 2D rozprostírání signálu a zjištění, že VSF-OFCDM systém překonává systémy využívající Orthogonal Division Frequency Multiplex (OFDM), nebo Code Division Multiple Access (CDMA). Dále byla navržena metoda 2D rozprostírání signálu v systému LTE, kde se též potvrdila její účinnost. Díky účinnějšímu potlačení vlivu rychlé variace přenosového kanálu v závislosti na frekvenci a čase, dosahuje systém VSF-OFCDM znatelně vyšší datové prostupnosti.

Multi-carrier systems suffer from several impairments, and communication system engineers use powerful signal processing tools to combat these impairments and keep up with the capacity/rate demands. Compressed sensing (CS) is one such tool that allows recovering any sparse signal, requiring only a few measurements in a domain that is incoherent with the domain of sparsity. Almost all signals of interest have some degree of sparsity, and in this work we utilize the sparsity of impairments in orthogonal frequency division multiplexing (OFDM) and its variants (i.e., orthogonal frequency division multiplexing access (OFDMA) and single-carrier frequency-division multiple access (SC-FDMA)) to combat them using CS. We start with the problem of peak-to-average power ratio (PAPR) reduction in OFDM. OFDM signals suffer from high PAPR and clipping is the simplest PAPR reduction scheme. However, clipping introduces inband distortions that result in compromised performance and hence needs to be mitigated at the receiver. Due to the high PAPR nature of the OFDM signal, only a few instances are clipped, these clipping distortions can be recovered at the receiver by employing CS. We then extend the proposed clipping recovery scheme to an interleaved OFDMA system. Interleaved OFDMA presents a special structure that result in only self-inflicted clipping distortions. In this work, we prove that distortions do not spread over multiple users (while utilizing interleaved carrier assignment in OFDMA) and construct a CS system that recovers the clipping distortions on each user. Finally, we address the problem of narrowband interference (NBI) in SC-FDMA. Unlike OFDM and OFDMA systems, SC-FDMA does not suffer from high PAPR, but (as the data is encoded in time domain) is seriously vulnerable to information loss owing to NBI. Utilizing the sparse nature of NBI (in frequency domain) we combat its effect on SC-FDMA system by CS recovery.

<p>High PAPR is one of the major challenges in using OFDM for communication, resulting in using inefficient linear amplifiers at the transmitter, or distortion if nonlinear amplifiers are used. Means of reducing PAPR in OFDM waveforms have been extensively researched, but those methods still have some need to avoid distortion. Using a pair of complementary sequences, an OFDM signal can be created that is constant in envelope, reducing the PAPR to unity and allowing the use of nonlinear amplifiers. Sequence pairs restricted to a particular alphabet are information-inefficient so instead a means of encoding onto a pair of complementary sequences is developed such that the length of the sequence increases linearly with the number of information symbols. The structure of this constant-modulus OFDM signal can also be used at the receiver to deal with channel fading and improve performance in noise. All of these concepts are extended to two-dimensional complementary sequence pairs for use in MIMO or broadcast applications using a uniform linear array while maintaining constant modulus on all antennas. This extension also includes exploration of additional structure imposed by the method of construction in multiple dimensions.</p>

The purpose of the project is to investigate the possibility of using modern machine learning to model nonlinear analog devices like the Power Amplifier (PA), and study the feasibility of using such models in wireless systems design. Orthogonal frequency division multiplexing (OFDM) is one of the most prominent modulation technique used in several standards like 802.11a,802.11n, 802.11ac and more. Telecommunication systems like LTE, LTE/Aand WiMAX are also based on OFDM. Nevertheless, OFDM system shows high peak to average power (PAPR) in time domain because it comprises of many subcarriers added via inverse fast Fourier transform(IFFT). HighPAPR results in an increased symbol error rate, while degrading the efficiency of the PA. Digital predistortion (DPD) still suffers from high symbol error rate (SER) and reduced PA efficiency, when there is an increase in peak back off(PBO). A receiver based nonlinearity distortion reduction approach can be justified by the fact that base stations have high computation power. A iterative-decision-feedback mitigation technique can be implemented as a receiver side compensation assuming memoryless PA nonlinearities. For successful distortion reduction the iterative-decision based technique required the knowledge of the transmitter PA. The author proposes to identify the nonlinear PA model using machine learning techniques like nonlinear regression and deep learning. The results show promising improvement in SER reduction with small PA model learning time. / Syftet med detta projekt är att undersöka möjligheterna att använda modernmaskininlärning för att beskriva ickelinjära analoga enheter såsom effektförstärkareoch att studera hur användbart det är att använda sådana modeller föratt designa trådlösa kommunikationssystem. OFDM (ortogonal frekvensmultiplex)är en av de vanligast förekommande modulationsteknikerna, som användsi standarder såsom 802.11a, 802.11n, 802.11ac and andra. Telekommunikationssystemsom LTE, LTE/A och WiMAX baseras också på OFDM. Dock resulterarOFDM i hög toppeffekt i förhållande till medeleffekten (hög PAPR) i tidsdomänen,eftersom signalen består av många delkanaler som summeras mha inversdiskret fouriertransform (IFFT). En hög PAPR resulterar i ökad symbolfelshaltoch försämrar effektiviteten hos effektförstärkaren. Digital predistortion (DPD)kan förbättra situationen men ger fortfarande hög symbolfelshalt och försämradförstärkareffektivitet, när man drar ned sändeffekten för undvika kvarvarandeickelineariteter. Att minska förvrängningen från ickelineariteterna vid mottagarenkan motiveras i system där basstationerna har hög beräkningsförmåga. Enmetod för att reducera förvrängningarna kan implementeras på mottagarsidan,baserad på iterativ beslutsåterkoppling, under antagandet om att sändarens effektförstärkarehar en minneslös ickelinearitet. För att störningsreduceringenska fungera väl, krävs god kunskap om sändarens effektförstärkare. Författarenföreslår att identifiera en ickelinjär modell för förstärkaren mha maskininlärningstekniker,såsom ickelinjär regression och djup inlärning. Resultaten visarlovande förbättringar av symbolfelshalten med en låg inlärningstid för förstärkarmodellen.