Search

Современные системы радиопередачи цифровых сигналов используют достаточно широкополосные трак-ты передачи, для которых характерно наличие частотно-селективных замираний, ведущих к межсимвольной интерференции (МСИ) принимаемых символов. Это приводит к заметному снижению помехоустойчивости связи вплоть до полного её срыва. Для борьбы с этим предлагается применять метод коррекции МСИ. Для этого перед каждым информационным символом излучается специально сформированный корректирующий символ. Его параметры регулируются таким образом, чтобы сумма с предыдущими «меша-ющими» символами была минимальной. Регулировка может осуществляться либо непосредственно амплитудно-фазовых параметров корректирующих символов («физическая» коррекция), либо с использованием особенностей метода QAM-модуляции. При QAM-модуляции для передачи информационных сим-волов уже производится амплитудная регулировка несущей с использованием QAM-модулятора передатчика. Подавая соответствующие логические коды, можно достаточно точно сформировать требующиеся корректирующие символы («логическая» коррекция). Исследована сравнительная эффективность обоих методов. Приведён пример реализации описанного метода «логической» коррекции., Higher data traffic via any transmission channels currently results in using signals increasingly broader in spec-trum and increasingly broader spectral bandwidth of transmission path. That said multipath propagation of radio waves is observed in many types of channels. Various beams come to the receiver along different paths and have varying levels and varying time delays. It results in signal power dissipation and adjacent symbols’ overlapping. Symbols interfere with each other and this causes their distortion and data transmission downgrade. To combat this phenomenon it is proposed to use the redundancy of transmitted symbols’ quantity compared to the quantity of initial information symbols. A special correction symbol is transmitted before each information symbol. It is readjusted so that to remove summation of all prior symbols or significantly reduce its level. The method is used in two-way transmission systems. Test sessions are conducted occasionally whereby the parameters of trans-mission channel and obstructive symbols that are added to the basic symbol are measured. The measured parameters are transmitted back to the sending station. Quite a sophisticated equipment is applied in the transmit-ter for tuning. This is how “physical” method of correction is implemented. Amplitude and phase of each correction symbol must be controlled. However, QAM (quadrature amplitude modulation) method enables to consider-ably simplify this procedure using “logical” correction method. Amplitude-phase control of transmitted signals is already performed in this method when transmitting information symbols. Therefore, it is quite enough to input such a binary code to the modulator instead of amplitude-phase control of the correction symbol, so that a symbol is formed at the modulator output and it is as similar as possible to the required type of correction symbol. The method loses to the "physical" correction method in terms of noise immunity, but discrepancies are slight in case of QAM-64, QAM-256 modulation. Implementation scheme for the proposed method is given. Computer-aided research results of the method efficiency are described

Приведены подходы при анализе искажений комплексных огибающих цифровых сигналов при распространении по спутниковым ионосферным радиолиниям, обусловливающих интерференционные помехи в дополнение к тепловым канальным шумам. Разработана методика оценки статистических характеристик интерференционных помех. Оценены вероятностные характеристики и асимптотические вероятности ошибочного приема рассматриваемых сигналов с увеличением частотной полосы с использованием вычисленных статистических характеристик для радиолиний Рчастотного диапазона., Models of signal propagation lines in transionosphere channels are presented. Modeling of the propagation of digital signals using these models during propagation with typical characteristics was carried out in order to evaluate the probabilistic characteristics of the reception of digital signals with the frequency band extension. The probabilistic characteristics and asymptotic probabilities of erroneous reception of the considered signals with increasing frequency band are estimated using the calculated statistical characteristics for radio links of the P-frequency range.

Системы, использующие фазовую модуляцию, широко используются при передаче цифровых сигналов. Для демодуляции принятых цифровых сигналов обычно применяется корреляционная обработка, заключающаяся в перемножении принятого сигнала на напряжение опорного генератора и интегрирование результатов перемножения на длительности символа. Проблемой является подстройка фазы генератора в приёмнике, используемого для детектирования. Наличие фазовой модуляции, переносящей передаваемую информацию, препятствует настройке генератора. Одним из способов устранения её влияния, является умножение частоты, однако при этом значительно возрастает уровень шумов за счёт появления комбинационных составляющих сигнала и шума. Такие схемы эффективны только в условиях большого отношения «сигнал/шум». Применяются схемы с петлёй Костаса, но они предлагались только для модуляции BPSK и QPSK. В статье предлагается модифицировать схему с петлёй Костаса для использования при модуляции 8-PSK. Приведены результаты компьютерного моделирования работы схемы., Phase modulation is in common use currently when transmitting digital signals. Thereby the generator should be used in the receiver for efficient detection, oscillations of which should be synphased with carrier frequency of in-coming signal. It is easy to do if the carrier is available. However, carrier suppression is often used for energy effi-ciency improvement. Then difficulties arise with phase lock for generator. These difficulties are resolved in various ways. Frequency multiplication of incoming signal is applied when using BPSK, QPSK and 8-PSK modulation types. Thereby phase surges due to transmitted information signal become multiple of 2π and do not affect phase lock process. However, when multiplying, signal and noise combination fluctuations arise which are particularly noticea-ble for 8-PSK modulation. Signal-to-noise ratio goes down and give rise to extra distortions. Costas circuit-based phase-lock devices have better performance properties. Two quadrature components of incoming signal are distin-guished there and phase lock is done through their co-processing. However, commonly known Costas circuits oper-ate only with BPSK and QPSK modulation types. The article proposes Costas circuit modification for 8-PSK modu-lation type. It is intended for differential modulation, when generator’s initial phase value can be any of the eight possible values. Operation concept lies in the fact that running sum of phases for generator and information signal equates to all eight values of possible phase and lock phase is done up to a proximate value. Efficiency of the pro-posed circuit was investigated. The results of computer-aided simulation are given.

Приведены модели линий распространения сигналов в лесах, основанные на их представлении в виде квазиоднородной среды с эффективной комплексной диэлектрической проницаемостью. Произведено моделирование распространения цифровых сигналов с использованием этих моделей при распространении в лесах с типичными характеристиками с целью оценивания вероятностных характеристик приема цифровых сигналов с расширением частотной полосы. Показано, что для класса цифровых сигналов с фазовой манипуляцией и с центральной частотой Pдиапазона с горизонтальной поляризацией энергетические потери по отношению к распространению в свободном пространстве с эквивалентным затуханием достигают 1.5-3.5 дБ., Models of signal propagation lines in forests based on their representation as a quasi-homogeneous medium with an effective complex permittivity are presented. Modeling of the propagation of digital signals using these models during propagation in forests with typical characteristics was carried out in order to evaluate the probabilistic characteristics of the reception of digital signals with the frequency band extension. It is shown that for the class of digital signals with phase shift keying and with the center frequency of the P-band with horizontal polarization, the energy losses in relation to propagation in free space with equivalent attenuation reach 1.5-3.5 dB.

Parkinson’s disease is a complex and heterogeneous condition, and there are many gaps in the medical community’s scientific and practical understanding of the disease. Closing these gaps relies on objective data about symptoms and signs, collected over long durations. Smartphones contain sensor devices which can be used to remotely capture behavioral signals. From these signals, computational algorithms can distill metrics of symptom severity and progression. This brief review introduces the main concepts of the discipline, addressing the experimental, hardware and software logistics, and computational analysis. The article finishes with an exploration of future prospects for the technology.

Kişilerin günlük yaşam kalitesini olumsuz bir şekilde etkilemekte olan uyku bozuklukları, artık toplumun büyük çoğunluğunun ortak problemi haline gelmiştir. Uyku ile ilgili en tehlikeli rahatsızlıkların başında, uykuda kendini gösteren ve hastaların uyku sırasında ani ölümlerine bile yol açabilen tıkayıcı uyku apnesi sendromu (OSAS) gelmektedir. Uyku evreleme işlemi bu tarz uyku rahatsızlıklarının teşhis ve tedavisinde büyük öneme sahip olmasının yanında uzmanlar için yorucu, zaman alıcı, deneyim isteyen ve uzmanlar arasında görüş ayrılıklarına yol açabilen bir uygulamadır. Bu ve benzeri nedenlerden dolayı otomatik uyku evreleme sistemleri tasarlanmaktadır. Uyku evreleme işlemi için elektroensefalogram (EEG) temel sinyal, elektrookülogram (EOG) ve elektromiyogram (EMG) sinyalleri ise yardımcı olarak kullanılmaktadır. Fakat bu sinyallerin hem elde edilmesi zor hem de yorumlanması subjektiftir. Bunun yanında hem kayıt edilmesi hem de değerlendirilmesi daha kolay olan elektrokardiyografi (EKG) sinyalinin kullanılmasının hem zaman hem de pratiklik açısından uzmanlara kazanç sağlayabileceği düşünülmektedir. Bu tez çalışmasında, uykunun EKG ile genel ilişkisinin sorgulanmasının yanında daha detaylı bir araştırma yapılarak OSAS, uyku evreleme ve uyku parametreleri gibi konuların kalp ritimleriyle olan bağlantıları incelenmiştir. Bu bağlantının ayrıntılı bir şekilde incelenebilmesi amacıyla üç ön ve üç ana çalışma olmak üzere toplam altı farklı çalışma yürütülmüştür. Çalışmalarda kullanılan uyku verileri Konya Necmettin Erbakan Üniversitesi Meram Tıp Fakültesi Uyku Laboratuvarından ve hazır veri tabanlarından alınmıştır. Bu tez çalışmasında, uyku EEG'sindeki EKG artifaktlarının temizlenmesi, EKG sinyalini temsil edebilen etkin morfolojik özelliklerin tespit edilmesi ve EKG sinyalinin yüzde değişim özellikleri kullanılarak uyku apne şiddetinin belirlenmesi olmak üzere üç farklı ön çalışma gerçekleştirilmiştir. Bu ön çalışmalardan ilkinin sonucunda, uyku evreleme amacıyla kullanılan temizlenmiş ve temizlenmemiş EEG sinyallerinin başarı oranları sırasıyla % 75.87 ve % 74.86 olarak elde edilmiştir. İkinci ön çalışma sonucunda ise, EKG sinyallerinden çıkarılan 15 farklı morfolojik özellikten 14'ünün kullanılması ile 12 aritminin sınıflandırılmasında başarı oranı % 1.5 iyileştirilerek % 98.96'ya ulaşılmıştır. Gerçekleştirilen son ön çalışma sonucunda, EKG özelliklerinin sadece yüzdesel değişimleri kullanılarak OSAS şiddeti % 84.38 oran ile tespit edilmiştir.Tez içerisinde gerçekleştirilen ana çalışmalar ise EKG parametreleri ile uyku parametreleri arasındaki ilişkinin belirlenmesi, EKG parametreleriyle uyku evreleme işleminin yapılması ve EKG parametreleri ile OSAS arasındaki ilişkinin tespiti şeklinde ifade edilebilir. İlk ana çalışma kapsamında, sadece EKG sinyalleri kullanılarak uyku evrelemenin belirleyici faktörleri olan ve EKG dışındaki diğer uyku sinyallerinde bulunan Arausal (AR), K-kompleksi (Kk), Uyku İğciği (Uİ), hızlı göz hareketleri (REM) ve yavaş göz hareketlerinin (SEM) tespit edilmesi çalışılmasında en yüksek performansla belirlenebilen uyku parametresi % 81.99 ile SEM olmuştur. İkinci ana çalışmada, OSAS rahatsızlığına sahip olan ve olmayan kişilerin gece boyunca kaydedilen EKG sinyalleri kullanılarak otomatik uyku evreleme sistemi tasarlanmış ve hasta ile sağlıklı veri grupları için sırasıyla % 78.08 ve % 87.11 başarı oranlarına ulaşılmıştır. Gerçekleştirilen son ana çalışmada ise, farklı OSAS seviyelerine sahip hastaların EKG sinyalleri kullanılarak beş farklı veri grubu oluşturulmuş ve en iyi sınıflama sonucu olarak % 97.20 elde edilmiştir.Tez kapsamında gerçekleştirilen tüm çalışmalardan elde edilen sonuçlar genel olarak değerlendirildiğinde, literatüre göre çok daha iyi sonuçların elde edildiği görülmüştür. Ayrıca literatür incelendiğinde, bu zamana kadar hiç denenmemiş sistem ve yöntemlerin bu tez çalışması içerisinde gerçekleştirilmiş olması da çalışmanın özgünlüğünü ortaya koymaktadır. Kısacası, elde edilen performans sonuçları, EKG'nin uyku ve OSAS ile olan ilişkisinin incelenmesi ve ayrıca uyku evreleme işleminin sadece EKG ile otomatik olarak gerçekleştirilmesi gibi bazı önemli konuların aydınlatılmasına yönelik gerçekleştirilen çalışmalar açısından bu tez çalışması literatürde önemli bir yere sahip olacaktır. Sleep disorders, which have a negative effect on the daily quality of life of individuals, have now become the common problem of a large majority of the society. Obstructive sleep apnea syndrome (OSAS), which manifests itself and may even cause sudden deaths in patients during sleep, ranks first among the most dangerous disorders related to sleep. Although sleep staging is of great importance in the diagnosis and treatment of such sleep disorders, it is a tiring and time consuming procedure that requires experience and may cause differences of opinion among specialists. For this and similar reasons, automatic sleep staging systems are designed. Electroencephalogram (EEG) signals are used as primary signals, and electrooculogram (EOG) and electromyogram (EMG) signals are used as secondary signals in the sleep staging process. However, these signals are difficult to obtain and their interpretation is subjective. Besides, it is considered that the use of electrocardiogram (ECG) signals, whose recording and evaluation is much easier, could provide benefit to specialists in terms of both time and practicality. In this thesis study, alongside the examination of the general relationship between sleep and ECG, the connections of topics such as OSAS, sleep staging, and sleep parameters to cardiac rhythms was investigated through a more detailed analysis. For a comprehensive examination of these connections, a total of six different studies were conducted consisting of three preliminary and three main studies. The sleep data used in the studies were taken from the sleep laboratory of Konya Necmettin Erbakan University Meram Faculty of Medicine and from ready databases.In this thesis study, three different preliminary studies were carried out that can be listed as the removal of ECG artefacts in the sleep EEG, determination of the effective morphological features that can represent the ECG signal, and the determination of the severity of the sleep apnea by using the percentage change features of ECG signals. As the result of the first preliminary study, the success rates of the cleaned and uncleaned EEG signals used for sleep staging were found as 75.87 % and 74.86 %, respectively. The results of the second preliminary study showed that the success rate of the classification of 12 arrhythmias by using 14 of the 15 different morphological features extracted from ECG signals reached 98.96 % with an improvement of 1.5 %. As the result of the last preliminary study, the severity of OSAS was determined at a rate of 84.38 % by using only the percentage changes in ECG features.The main studies conducted within the thesis can be stated as the determination of the relationship between ECG parameters and sleep parameters, sleep staging through sleep parameters, and the determination of the relationship between ECG parameters and OSAS. Within the scope of the first main study conducted to determine Arousal (AR), K-complex (Kc), sleep spindle (SS), rapid eye movements (REM), and slow eye movements (SEM), which are the determining factors of sleep staging and exist in sleep signals other than ECG, the sleep parameter that could be determined with the highest performance was SEM with a rate of 81.99 %. In the second main study, an automatic sleep staging system was designed using the ECG signals from individuals with and without OSAS recorded throughout the night, and success rates of 78.08 % and 87.11 % were achieved for ill and healthy data groups, respectively. In the final main study, five different data groups were created by using the ECG signals from patients with different degrees of OSAS severity, and the optimal classification result was achieved as 97.20 %.A general evaluation of the results obtained from all the analyses conducted within the scope of the thesis study show that the present study yielded much better results compared to the literature. In addition, the performance of systems and methods that were not previously conducted in the literature for the first time within this thesis study presents the authenticity of the study. In brief, this thesis study will occupy a prominent place in the literature with the performance results obtained in the studies, the investigation of the relationship of ECG with sleep and OSAS, and also in terms of the procedures conducted to shed light on certain important issues such as the performance of automatic sleep staging through ECG alone. 147

El mantenimiento de la vía se ha convertido en los últimos años en un tema verdaderamente relevante para la ingeniería ferroviaria debido a la necesidad de conseguir mejores niveles de calidad a menores costes. Para ello, no hay más remedio que incorporar a las filosofías y procedimientos tradicionales del mantenimiento y la auscultación de la vía nuevas técnicas, que en muchos casos están ampliamente desarrolladas en otros campos de la ingeniería. Así, este trabajo profundiza en las técnicas de adquisición y procesado de datos para su aplicación a la auscultación de la vía. Para ello, se han realizado varias pruebas en la red de Metro de Valencia, registrando las aceleraciones producidas en las cajas de grasa durante la circulación de los vehículos por las vías. Dichas aceleraciones se han registrado y analizado de diferentes maneras, variando la frecuencia de muestreo, la frecuencia de filtrado, el emplazamiento de los acelerómetros y los diferentes parámetros que definen las representaciones tiempo-frecuencia. Los resultados muestran los valores óptimos de los parámetros que determinan la auscultación dinámica y el mejor emplazamiento para los acelerómetros. Además, mediante el análisis espectral y los diagramas tiempo-frecuencia es posible identificar y clasificar los diferentes defectos, puntos singulares y modos de vibración de la vía. De este modo, pueden sentarse las bases para la aplicación de técnicas de procesado digital de imágenes al mantenimiento de la vía., Salvador Zuriaga, P. (2014). Desarrollo de técnicas avanzadas en la auscultación dinámica para la optimización del mantenimiento de vías férreas [Tesis doctoral]. Editorial Universitat Politècnica de València. doi:10.4995/Thesis/10251/39003., Alfresco

Los registros de pozo presentan gran variedad de información que permite caracterizar propiedades físicas y químicas de las rocas así como de los fluidos contenidos entre ellas. Los muestreos geofísicos en sitio caracterizan en detalle parámetros físicos de las formaciones que permiten determinar la resistividad, densidad volumétrica, litología, permeabilidad y porosidad entre otros. Dentro de los avances en predicción de fracturas naturales se han desarrollado modelos soportados por procesos dependientes de la dimensión fractal de imágenes resistivas. Sin embargo, no siempre se cuenta con la información asociada a dichas imágenes. Esta investigación tiene como base la información de registros geofísicos de pozo y parte de la hipótesis que existen relaciones significativas entre múltiples registros con el objeto de caracterizar los valores de dimensión fractal. Se implementa un modelo de predicción utilizando técnicas de minería de datos y bajo los lineamientos de la metodología CRISPDM. El modelo permite de una forma adaptativa el procesamiento de información proveniente de los registros de pozo con el objeto de identificar cuáles de estos registros presentan relaciones significativas para la generación de patrones que inciden en la predicción de valores de dimensión fractal. El modelo es validado sobre un caso de estudio con información de registros de pozo realizados en la cuenca del Catatumbo, Norte de Santander, Colombia. En los capítulos asociados se presenta el análisis de información, ejecución y evaluación del modelo. El análisis de resultados de ejecución para el modelo de predicción se basa en muestreos aleatorios de valores de registros múltiples en secciones transversales y es evaluado sobre valores aleatorios de dimensión fractal. Como producto del proceso de clasificación se identificaron tres conjuntos de orígenes de datos relevantes como son registros (logs) base, registros (logs) complementarios y registros (logs) no válidos para la generación de valores de dimensión fractal. La predicción asociada al caso de estudio evidenció una respuesta altamente cercana hacia los valores esperados con una tasa de error baja donde el coeficiente de determinación (R2) resultante es de 0.95, la media absoluta del error igual a 0.003 y el error de la raíz cuadrada de la media (RMSE) es igual a 0.004. La metodología propuesta ofrece una alternativa para generar e introducir valores asociados a la dimensión fractal de imágenes resistivas en escenarios donde se cuenta con información de diversos registros geofísicos de pozo pero donde la información asociada con la dimensión fractal es inexistente. Abstract. Well logs have a great variety of information to characterize physical and chemical properties of rock formations and its fluid content among them. Geophysical surveys characterize detailed physical parameters of rock formations in order to determine the resistivity, bulk density, lithology, porosity and permeability, inter alia. Among the advances made on prediction of natural fractures, there have been developed different models supported by processes that depend on fractal dimension of resistive images. However, the information associated to such resistivity image logs is not always available. The current research is based on information from geophysical well logs, and also upon the hypothesis that is possible to identify significant relationships between multiple records in order to characterize the fractal dimension values. A prediction model is implemented through a data mining technique under the guidelines of CRISP-DM methodology. The model allows an adaptively way of information processing based on well logs in order to identify which of these records show significant relationships to generate prediction patterns to estimate fractal dimension values. The model is validated over a study case with well-log information gathered on-site at the Catatumbo basin, Norte de Santander, Colombia. In the study case chapters it is also presented data analysis, model execution and evaluation. The analysis of the execution results for the prediction model is based on random samples of multiple values in cross section and is evaluated upon random fractal dimension values. As a product of classification process three groups of relevant data sources were identified, such as: base logs, complementary logs and invalid logs for the generation of fractal dimension values. The prediction associated to the case study showed a very accurate response, in regards to the expected result values, with a low error rate where the resulting R2 is 0.95, the mean absolute error is equal to 0.003 and the root mean squared error is 0.004. The proposed methodology is an alternative to generate and input values associated to the fractal dimension of resistive images in scenarios where well geophysical logs are available, but information regarding values of fractal dimensions is inexistent. Maestría

В работе рассматриваются два метода определения вида манипуляции. Разрабатывается алгоритм определения метода манипуляции Ñ†Ð¸Ñ„Ñ€Ð¾Ð²Ñ‹Ñ ÑÐ¸Ð³Ð½Ð°Ð»Ð¾Ð² с квадратурными методами манипуляции. Производится программная реализация данного алгоритма в среде Matlab. Исследуется зависимость вероятности ошибки от отношения сигнал/шум. ÐÐ°Ñ Ð¾Ð´Ð¸Ñ‚ÑÑ вычислительная сложность разработанного метода.