[pt] Esta dissertação enfoca a combinação de duas conhecidas e eficazes tecnologias usadas em muitas redes de comunicação atuais, as técnicas de transmissão Orthogonal Frequency Division Multiplexing (OFDM) e Multiple-Input Multiple-Output (MIMO). Sistemas MIMO-OFDM unem as vantagens conhecidas da comunicação MIMO, como altas taxas de dados e ampliação da confiabilidade do link, com as características marcantes da transmissão OFDM, como comunicação livre de interferência entre símbolos e a equalização simples no domínio da frequência. Este trabalho apresenta em detalhes a modelagem de sinais do sistema MIMO-OFDM e o processamento adequado no receptor para realizar detecções desacopladas dos sub-símbolos OFDM. Além disso, estratégias de seleção de antenas são propostas para aumentar o desempenho do sistema, explorando cenários em que o transmissor ou receptor está equipado com mais antenas do que o número de cadeias de radiofrequência (RF). Esses procedimentos são baseados em métricas específicas propostas para cada cenário. A avaliação de desempenho inclui sistemas equipados com técnicas de equalização linear e sistemas com pré-codificadores, como Zero-Forcing (ZF) e Minimum Mean Square Error (MMSE). Os resultados de simulação evidenciam que as estratégias de seleção da antena resultam em uma melhoria significativa no desempenho do sistema. / [en] This thesis focuses on the combination of two well-known effective technologies used in many of today s communication networks, Orthogonal Frequency Division Multiplexing (OFDM) and Multiple-Input Multiple- Output (MIMO). It unites the well-known advantages of MIMO communication, such as high data rates and improved link reliability to the characteristics of OFDM, known as intersymbol-interference-free communication and simple frequency-domain equalization. This work presents in detail the signal modelling of MIMO-OFDM system and the adequate processing at the receiver to perform decoupled detections per OFDM subsymbol. In addition, antenna selection strategies are proposed to increase system performance by exploiting some scenarios when the transmitter or receiver is equipped with more antennas than the number of radiofrequency (RF) chains. These procedures are based on the specific metric for the current scenario. The performance of the system is evaluated employing linear equalization or precoding techniques, such as Zero-Forcing (ZF) and Minimum Mean Square Error (MMSE). Simulation results evidence that antenna selection strategies result in significant system performance improvement.

[pt] Esta dissertação enfoca o enlace direto de sistemas MIMO multiusuário com pré-codificação onde a estação base e os terminais dos usuários possuem múltiplas antenas mas transmitem e recebem, respectivamente, símbolos de informação através de subconjuntos selecionados de seus conjuntos de antenas. O trabalho considera sistemas que utilizam técnicas de précodificação linear como Zero Forcing (ZF) e Minimum Mean Square Error (MMSE). Expressões gerais que descrevem os sistemas e relacionam a energia gasta na transmissão com a energia disponível para a detecção em cada usuário são apresentadas. Com base nestas relações, um procedimento para seleção de antenas na transmissão é proposto visando a minimização da probabilidade de erro. Um algoritmo de busca não exaustiva denominado ITES (Iterative Search) foi desenvolvido e testado e mostrou-se capaz de, com apenas uma pequena fração do esforço computacional, fornecer um desempenho próximo ao da seleção ótima, que demanda uma busca exaustiva. A seleção de antenas na recepção é também efetuada usando um critério de otimização semelhante. O caso geral da seleção conjunta de antenas na transmissão e na recepção contempla a combinação de ambas estratégias, resultando na redução da complexidade tanto na estação base, quanto nos terminais dos usuários. Os resultados de desempenho em termos da taxa de erro de bit, obtidos por meio de simulações e abordagem semianalítica, são apresentados para diferentes cenários. / [en] This thesis focuses on the downlink of a multiuser multiple-input multiple-output (MU-MIMO) systems where the Base Station (BS) and the users stations (UEs) transmit and receive information symbols, respectively, by selected subset of their antennas. The performance of the system is evaluated employing linear precoding techniques as Zero Forcing (ZF) and Minimum Mean Square Error (MMSE). A general model to describe the system and expressions that relate the energy spent in transmission with the energy available for detection at each user are presented. A transmit antenna selection procedure is proposed aiming at the minimization of the detection error probability. A suboptimal search algorithm, called ITES (Iterative Search), able to deliver a performance close to the one resulting from the optimal exhaustive search selection is also proposed. The receive antenna selection is also performed using a similar optimization criterion. Joint antennas selection at the transmitter and receiver contemplates the efficient combination of both strategies, leading to a complexity reduction in BS and UEs. BER performance results, obtained via simulation and semi-analytical approaches, are presented for different scenarios.

Abstract This thesis concentrates on the design and evaluation of spatial user multiplexing methods via linear transmit-receive processing for wireless cellular multi-user multiple-input multiple-output (MIMO) communication systems operating in the time-division duplexing (TDD) mode. The main focus is on the acquisition of effective channel state information (CSI) that facilitates decentralized processing so that the network nodes – base stations (BS) and user terminals (UT), each employing an arbitrary number of antenna elements – are able to locally participate in the network adaptation. The proposed methods rely on the uplink-downlink channel reciprocity and spatially precoded over-the-air pilot signaling. Considering (single-cell) multi-user MIMO systems, coordinated zero-forcing transmit-receive processing schemes for the uplink (UL) are proposed. The BS computes the transmission parameters in a centralized manner and employs downlink (DL) pilot signals to convey the information of the beamformers to be used by the UTs. When coexisting with the DL zero-forcing, the precoded DL demodulation pilots can be reused for UL beam allocation, and the precoded UL demodulation pilots are reused in turn for partial channel sounding (CS). As a result, only the precoded pilot symbols are needed in both UL and DL. Moreover, a concept for reducing the number of the required orthogonal UL CS pilot resources is presented. Based on their DL channel knowledge, the multi-antenna UTs form fewer pilot beams by spatial precoding than conventionally needed when transmitting antenna-specific pilots. In the context of DL zero-forcing, when taking into account the CSI estimation error at the BS, the overhead reduction turns out to improve robustness and increase the average system capacity. Considering multi-cell multi-user MIMO systems, decentralized coordinated DL beamforming strategies based on weighted sum rate (WSR) maximization are proposed. An optimization framework where the WSR maximization is carried out via weighted sum mean-squared-error minimization is utilized, and the approach is generalized by employing antenna-specific transmit power constraints. The iterative processing consists of optimization steps that are run locally by the BSs. In one novel strategy, the coordinating cells update their transmit precoders and receivers one cell at a time, which guarantees monotonic convergence of the network-wide problem. The strategy employs separate uplink CS and busy burst pilot signaling to reveal the effective channels of the UTs to the neighboring BSs. In another novel strategy, the monotonic convergence is sacrificed to devise a faster scheme where the BSs are allowed to optimize their variables in parallel based on just the CS responses and additional low-rate backhaul information exchange. The numerical results demonstrate that WSR maximization has the desirable property that spatial user scheduling is carried out implicitly. Finally, methods for UL CS overhead reduction are presented, and the effect of CSI uncertainty is addressed. / Tiivistelmä Tämä väitöskirja keskittyy lineaarisella lähetys- ja vastaanottoprosessoinnilla toteutettavien tilajakomonikäyttömenetelmien suunnitteluun ja arviointiin langattomissa moniantennisissa solukkoverkoissa, jotka hyödyntävät aikajakodupleksointia (TDD). Erityisesti tarkastellaan efektiivisen kanavatiedon hankintaa, joka mahdollistaa hajautetun prosessoinnin siten että verkkoelementit – tukiasemat ja terminaalit, jotka kukin hyödyntävät useaa antennielementtiä – voivat osallistua paikallisesti verkon adaptaatioon. Esitetyt menetelmät perustuvat ylä- ja alalinkin kanavien resiprookkisuuteen ja tilatasossa esikoodattuun opetus- eli pilottisignalointiin ilmarajapinnan yli. Yksisoluisille monikäyttäjä- ja moniantennijärjestelmille esitetään ylälinkin koordinoituja nollaanpakottavia lähetys- ja vastaanottomenetelmiä. Tukiasema laskee lähetysparametrit keskitetysti ja käyttää pilottisignaaleja kertomaan millaista lähetyskeilanmuodostusta terminaalien tulee käyttää. Alalinkin nollaanpakotuksen yhteydessä esikoodattuja demodulaatiopilotteja voidaan uudelleenkäyttää ylälinkin lähetyskeilojen allokointiin, ja esikoodattuja ylälinkin demodulaatiopilotteja uudelleenkäytetään puolestaan osittaiseen kanavan luotaukseen (sounding). Näin ollen molempiin suuntiin tarvitaan vain esikoodatut pilotit. Lisäksi työssä esitetään menetelmä ylälinkin luotauspilottiresurssitarpeen vähentämiseksi. Kanavatietoon perustuen moniantenniset terminaalit muodostavat tilatasossa esikoodattuja pilottilähetyskeiloja, joita tarvitaan vähemmän kuin perinteisiä antennikohtaisia pilotteja. Kun otetaan huomioon kanavanestimointivirhe tukiasemassa, resurssiensäästömenetelmä parantaa häiriösietoisuutta ja nostaa järjestelmän keskimääräistä kapasiteettia alalinkin nollaanpakotuksen yhteydessä. Monisoluisille monikäyttäjä- ja moniantennijärjestelmille esitetään hajautettuja koordinoituja alalinkin keilanmuodostusstrategioita, jotka perustuvat painotetun summadatanopeuden (WSR) maksimointiin. Valitussa optimointikehyksessä WSR:n maksimointi toteutetaan painotetun summaneliövirheen minimoinnin kautta, ja työssä menettelytapa yleistetään antennikohtaisten lähetystehorajoitusten tapaukseen. Iteratiivinen prosessointi koostuu optimointiaskelista, jotka tukiasemat paikallisesti suorittavat. Yhdessä esitetyssä strategiassa yhteistoiminnalliset solut päivittävät lähettimensä ja vastaanottimensa yksi solu kerrallaan, mikä takaa verkonlaajuisen ongelmanratkaisun monotonisen konvergenssin. Tämä strategia käyttää erillisiä ylälinkin luotaussignaaleja sekä varattu-signaaleja ilmaistakseen terminaalien efektiiviset kanavat naapuritukiasemille. Toisessa strategiassa monotoninen konvergenssi uhrataan ja kehitetään nopeammin adaptoituva menetelmä, jossa tukiasemat saavat optimoida muuttujansa rinnakkain, perustuen vain luotaussignaaleihin ja tukiasemien väliseen informaationvaihtoon. Numeeriset tulokset osoittavat, että WSR:n maksimointi toteuttaa aktiivisten käyttäjien valinnan tilatasossa implisiittisesti. Lopuksi esitetään menetelmiä luotauspilottiresurssitarpeen vähentämiseksi ja käsitellään kanavatiedon epävarmuuden vaikutusta.

Abstract This thesis considers data transmission in a full-duplex (FD) multiuser multiple-input multiple-output (MU-MIMO) system, where a FD capable base station (BS) bidirectionally communicates with multiple half-duplex (HD) users in downlink (DL) and uplink (UL) channels using the same radio resources. The main challenge in FD communications is how to deal with the self-interference (SI) between transmit and receive antennas at the BS. The work carried out in the thesis is motivated by recent advanced techniques in hardware design demonstrating that the SI can be suppressed to a degree that possibly allows for FD transmission in cellular networks. In particular, this thesis attempts to explore the potential gains in terms of the spectral efficiency (SE) and energy efficiency (EE) that can be brought by the FD MU-MIMO model. As the first of its kinds, the thesis aims to present a solid mathematical framework and report interesting results that foster research on wireless communications in general and FD communications in particular. For the FD system of interest the major challenge is due to the SI and co-channel interference from users in the UL channel to the ones in the DL channel, resulting in the coupling between the two channels. As a result we are concerned with the problem of joint transmit processing design to maximize the SE and EE subject to certain power constraints. Since the design problems are natually non-convex, it is difficult to find the globally optimal solutions or even when possible it is not practically appealing. Our contributions to solving these design problems are on the development of several iterative algorithms that can obtain locally optimal solutions. The proposed algorithms are built upon a framework of local optimization strategies such as the sequential parametric convex approximation and the Frank-Wolfe methods. In special cases closed-form designs are also presented. The reported results show that when the SI is sufficiently suppressed the considered FD MU-MIMO system with the proposed SE designs achieves a significantly better SE but consumes more energy, compared to the HD counterpart. In terms of EE the proposed EE scheme is superior to the proposed SE design. Moreover, in the low transmit power region, the EE design achieves a worse EE than the HD system but a better one in the high trasmit power regime when the SI power is low. / Tiivistelmä Tämä väitöskirja käsittelee datansiirtoa samanaikaisesti kaksisuuntaisessa (full-duplex, FD) usean käyttäjän moniantennijärjestelmässä (MU-MIMO), jossa FD-kykyinen tukiasema on yhtä aikaa yhteydessä vuorosuuntaisten (half-duplex, HD) käyttäjien kanssa laskevalla (DL) ja nousevalla (UL) siirtotiellä käyttäen samoja radioresursseja. FD-kommunikaation suurin haaste liittyy lähetys- ja vastaanottoantennien välisen omahäiriön (SI) hallintaan. Tässä työssä hyödynnetään tuoreita tutkimustuloksia, joissa edistyneillä häiriönvaimennustekniikoilla on kyetty vaimentamaan omahäiriö tasolle, jolla FD-lähetys solukkoverkoissa on toteutuskelpoista. Tässä työssä tutkitaan etenkin mahdollisia FD MU-MIMO –järjestelmän tuomia suorituskykyparannuksia spektrinkäytön tehokkuudessa (SE) ja energiatehokkuudessa (EE). Väitöskirjalla on uutuusarvoa matemaattisessa suorituskykyarvioinnissa ja työn mielenkiintoiset tulokset edistävät jatkotutkimusta aiheen ympärillä. Tutkittavan FD-järjestelmän merkittävänä haasteena on omahäiriön ja muiden käyttäjien siirtosuuntien välisen samankanavan häiriön yhteisvaikutus, jonka johdosta siirtosuunnat kytkeytyvät toisiinsa. Tämä johtaa lähetysprosessoinnin yhteisoptimointiin, jossa spektri- ja energiatehokkuus pyritään maksimoimaan määritetyillä tehorajoituksilla. Nämä suunnitteluongelmat eivät ole luonteeltaan konvekseja, joten niihin on vaikeaa löytää globaalisti optimaalisia ratkaisuja ja vaikka onnistuisikin niin ne eivät yleensä ole käytännöllisiä. Työssä esitetään useita iteratiivisia algoritmejä, joilla saavutetaan paikallisesti optimaalisia ratkaisuja. Ehdotetut algoritmit pohjautuvat paikallisten optimointistrategioiden viitekehykseen, jossa käytetään esimerkiksi peräkkäistä parametristä konveksiapproksimaatiota ja Frank-Wolfe –menetelmiä. Erityistapauksissa suljetun muodon ratkaisut on myös esitetty. Raportoidut tulokset osoittavat, että omahäiriön ollessa riittävästi vaimennettu mallinnetulla järjestelmällä saavutetaan spektrinkäytön optimointimielessä huomattavaa etua HD-verrokkiin lisääntyneen energian kulutuksen kustannuksella. Energiatehokkuuden optimointiin pohjautuvalla strategialla puolestaan päästään suurempiin suorituskykyetuihin. Pienillä lähetystehoilla energiatehokkuus voi kuitenkin olla HD-järjestelmää alempi, mutta vastaavasti suurten lähetystehojen alueella tilanne on päinvastainen kunhan omahäiriön teho on tarpeeksi alhainen.

Les réseaux radio actuelles utilisent le spectre inefficacement, car une bande de fréquence est allouée de façon permanente à une technologie spécifique. Vu que le spectre est une ressource limitée, cette attribution statique ne pourra bientôt plus combler les besoins des systèmes de transmission qui ne cessent de croître. On peut toutefois optimiser l'utilisation du spectre en permettant des transmissions secondaires (SU) dans les espaces libres du primaire (PU). Cette vision constitue l'objectif principal de la radio cognitive. Nous proposons d'évaluer les stratégies de transmission pour la coexistence des systèmes primaires (PU) et SU dans les mêmes réseaux. Plus concrètement, nous nous focalisons sur un scénario spatial interweave en émettant les signaux SU dans les espaces vides du PU à l'aide d'un précodeur linéaire. Néanmoins, ce précodage nécessite une connaissance a priori des canaux interférents. L'échange d'informations entre le PU et le SU étant proscrit, nous exploitons l'hypothèse de la réciprocité du canal. Cette hypothèse compense l'absence de coopération, mais elle n'est pas si évidente à exploiter en pratique à cause des perturbations des circuits radio fréquence. Nous suggérons de compenser ces perturbations par des méthodes de calibration relative. Nous proposons ensuite une implémentation temps-réel des solutions sur une plateforme LTE. Pour finir, nous généralisons l'approche RC à un système de transmission multi-utilisateurs, à travers une combinaison des techniques RC et massive MIMO, cette approche constitue s’établit comme une solution à la progression exponentielle du trafic. / The accelerated evolution of wireless transmission in recent years has dramatically increased the spectrum overcrowding. Indeed, the spectrum is inefficiently used in the conventional networks, since a frequency band is statically allocated to a specific technology called primary (PU). Whereas the radio spectrum is limited, this static frequency allocation will no longer be able to meet the increasing needs of bandwidth. However, the spectrum can be optimally used in enabling secondary (SU) transmissions, provided the latters do not harm the PU. This opportunistic vision of wireless transmissions is the main aim of Cognitive Radio (CR). CR enables smart use of wireless resources and is a key ingredient to perform high spectral efficiency. We focus on a spatial interweave (SIW) CR scenario which exploits the spatial white spaces to enable SU transmissions. The latter forms spatial beams using precoders, so that there is no interference towards the primary. Nevertheless, this precoding requires acquisition of the crosslink channel. However, due to the lack of cooperation between PU and SU, we acquire the channel thanks to channel reciprocity. Furthermore, the practical use of the reciprocity is not as straightforward as in theory, because it is is jeopardized by the nonreciprocal radio frequency front-ends. These perturbations are compensated in our study by relative calibration algorithms. Subsequently, we propose an implementation of our solutions in a real-time LTE platform. Eventually, we extend the CR model to a MU system in suggesting a combination of SIW and massive MIMO techniques. This scheme is an interesting candidate to overcome the exponential traffic growth.

Multiple-input, multiple-output (MIMO) techniques have been proposed to meet the needs for higher data rates and lower delays in future wireless communication systems. The downlink capacity of multiuser MIMO systems is achieved when the system transmits to several users simultaneously. Frequently, many more users request service than the transmitter can simultaneously support. Thus, the transmitter requires a scheduling algorithm for the users, which must balance the goals of increasing throughput, reducing multiuser interference, lowering delays, ensuring fairness and quality of service (QoS), etc. In this thesis, we investigate the application of genetic algorithms (GAs) to perform scheduling in multiuser MIMO systems. GAs are a fast, suboptimal, low-complexity method of solving optimization problems, such as the maximization of a scheduling metric, and can handle arbitrary functions and QoS constraints. We first examine a system that transmits using capacity-achieving dirty paper coding (DPC). Our proposed GA structure both selects users and determines their encoding order for DPC, which affects the rates they receive. Our GA can also schedule users independently on different carriers of a multi-carrier system. We demonstrate that the GA performance is close to that of an optimal exhaustive search, but at a greatly reduced complexity. We further show that the GA convergence time can be significantly reduced by tuning the values of its parameters. While DPC is capacity-achieving, it is also very complex. Thus, we also investigate GA scheduling with two linear precoding schemes, block diagonalization and successive zero-forcing. We compare the complexity and performance of the GA with "greedy" scheduling algorithms, and find the GA is more complex, but performs better at higher signal-to-noise ratios (SNRs) and smaller user pool sizes. Both algorithms are near-optimal, yet much less complex than an exhaustive search. We also propose hybrid greedy-genetic algorithms to gain benefits from both types of algorithms. Lastly, we propose an improved method of optimizing the transmit covariance matrices for successive zero-forcing. Our algorithm significantly improves upon the performance of the existing method at medium to high SNRs, and, unlike the existing method, can maximize a weighted sum rate, which is important for fairness and QoS considerations. / Communications

No description available.