Surface-generalized ambient noise in a shallow ocean waveguide with a sediment layer possessing a specific class of density and sound speed distributions capable of describing a realistic seabed environment is considered in this analysis. This class of non-uniform sediment layer has the density and sound speed distributions varying with respect to depth as a genearlized-exponential and an inverse-square function, respectively. The study invokes a formulation developed by Kuperman and Ingenito for surface noise generation, in conjunction with the analytical solutions for the Helmholtz equation corresponding to the sediment layer, to arrive at an analytical expression convenient for numerical implementation. The intensity and spatial correlation of the noise sound field are analyzed with respect to the variations of the system parameters, including frequency, sediment layer thickness, sound speed gradient, with emphasis on the effects of sediment properties on the ambient noise field. The results have demonstrated that the intensity of the noise field is relatively sensitive to the variations of the paramters, while the spatial correlation is not, suggesting that the energy distribution, rather than the spatial structure, of the noise field is susceptible to the environmental variation.

This thesis considers resource management issues in wireless sensor networks (WSNs), wireless local area networks (WLANs), and cognitive radio (CR) networks. Since energy is a critical resource in WSNs, we consider energy minimization techniques based on explicit node cooperation and distributed source coding (DSC). The explicit node cooperation based on space time block codes (STBC) improves energy efficiency of WSNs, by reducing the energy consumption per bit of each sensor node. The DSC on the other hand exploits the spatial correlation in WSNs, and thus reduces the data generated in a WSN. For the purpose of our analysis, we model the spatial correlation according to a linear Gauss-Markov model. Through our numerical results, we observe that the node cooperation combined with DSC can improve energy efficiency for many cases of interest. A unique aspect of our work is we obtain important structural results using the concepts from monotone comparative statics. These structural results provide insights into the general design of WSNs. Through our numerical results, we also demonstrate that, the cooperation based transmission can achieve better mutual information (MI)-energy tradeoff than the non-cooperation based transmission scheme. From the perspective of WLANs, we propose a price based approach to regulate the channel occupancy of low rate users, which is known to be the primary cause for low overall throughput in WLANs. Owing to the decentralized nature of WLANs we use non-cooperative game theory as a tool for analysis. Specifically, we use supermodular game theory. Through our analysis, we show that an increase in price leads to an increase in rate of WLAN users. We also prove that the best response dynamics indeed converge to the Nash equilibrium of the underlying non-cooperative game. Through our numerical results, we demonstrate that by proper tuning of the price, the proposed price based approach can lead to an improvement in overall throughput of a WLAN. Finally from the perspective of CR networks, we consider the impact of number of channels captured by a secondary user on its transmission control protocol (TCP) throughput. From our simulation results it was found that, there exists a definite optimal number of channels a secondary user needs to capture, to maximize its TCP throughput.

This thesis considers resource management issues in wireless sensor networks (WSNs), wireless local area networks (WLANs), and cognitive radio (CR) networks. Since energy is a critical resource in WSNs, we consider energy minimization techniques based on explicit node cooperation and distributed source coding (DSC). The explicit node cooperation based on space time block codes (STBC) improves energy efficiency of WSNs, by reducing the energy consumption per bit of each sensor node. The DSC on the other hand exploits the spatial correlation in WSNs, and thus reduces the data generated in a WSN. For the purpose of our analysis, we model the spatial correlation according to a linear Gauss-Markov model. Through our numerical results, we observe that the node cooperation combined with DSC can improve energy efficiency for many cases of interest. A unique aspect of our work is we obtain important structural results using the concepts from monotone comparative statics. These structural results provide insights into the general design of WSNs. Through our numerical results, we also demonstrate that, the cooperation based transmission can achieve better mutual information (MI)-energy tradeoff than the non-cooperation based transmission scheme. From the perspective of WLANs, we propose a price based approach to regulate the channel occupancy of low rate users, which is known to be the primary cause for low overall throughput in WLANs. Owing to the decentralized nature of WLANs we use non-cooperative game theory as a tool for analysis. Specifically, we use supermodular game theory. Through our analysis, we show that an increase in price leads to an increase in rate of WLAN users. We also prove that the best response dynamics indeed converge to the Nash equilibrium of the underlying non-cooperative game. Through our numerical results, we demonstrate that by proper tuning of the price, the proposed price based approach can lead to an improvement in overall throughput of a WLAN. Finally from the perspective of CR networks, we consider the impact of number of channels captured by a secondary user on its transmission control protocol (TCP) throughput. From our simulation results it was found that, there exists a definite optimal number of channels a secondary user needs to capture, to maximize its TCP throughput. / Applied Science, Faculty of / Electrical and Computer Engineering, Department of / Graduate

It is often important to detect an increase in the frequency of some event. Particular attention is given to medical events such as mortality or the incidence of a given disease, infection or birth defect. Observations are regularly taken in which either an incidence occurs or one does not. This dissertation contains the result of an investigation of prospective monitoring techniques in two distinct surveillance situations. In the first situation, the observations are assumed to be the results of independent Bernoulli trials. Some have suggested adapting the scan statistic to monitor such rates and detect a rate increase as soon as possible after it occurs. Other methods could be used in prospective surveillance, such as the Bernoulli cumulative sum (CUSUM) technique. Issues involved in selecting parameters for the scan statistic and CUSUM methods are discussed, and a method for computing the expected number of observations needed for the scan statistic method to signal a rate increase is given. A comparison of these methods shows that the Bernoulli CUSUM method tends to be more effective in detecting increases in the rate. In the second situation, the incidence information is available at multiple locations. In this case the individual sites often report a count of incidences on a regularly scheduled basis. It is assumed that the counts are Poisson random variables which are independent over time, but the counts at any given time are possibly correlated between regions. Multivariate techniques have been suggested for this situation, but many of these approaches have shortcomings which have been demonstrated in the quality control literature. In an attempt to remedy some of these shortcomings, a new control chart is recommended based on a multivariate exponentially weighted moving average. The average run-length performance of this chart is compared with that of the existing methods. / Ph. D.

No description available.

Nowadays, identification of fake data is an elaboratechallenge that calls for the use of machine learning techniques.This is due to the huge amount of data and its complexity makesthe differences indistinguishable even for the trained eye. In thisproject we use the MATLAB wlanTGnChannel System objectto simulate multipath fading channels that are comparable toreal impulse response measurements made by Ericsson AB of anindoor8×8MIMO (Multiple Input Multiple Output) system.We use an SVM classifier to compare the eigenvalues of theircorrelation covariance matrices, obtaining an accuracy of 84%.Comparing their power delay profiles (PDPs) happens to bea classification task of low complexity due to time resolutionlimitation in the real measurements. We suggest that the proposedMATLAB model strongly differs from the real data we have beenprovided with. / Nu för tiden så är identifiering av fejkad data en svår utmaning som ofta kräver maskininlärningstekniker. Detta beror på den stora mängden data och att komplexiteten i datat gör att skillnaderna kan vara svår att se även för ett tränat öga. I det här projektet använder vi oss av MATLABs systemobjekt wlanTGnChannel för att simulera flervägs fädningskanaler som kan jämföras med riktiga impulssvarsmätningar gjorda av Ericsson AB av ett innomhus 8 X 8 MIMO(Multiple Input Multiple Output) system. Vi använde en SVM (stödvektormaskins) klassificerare för att jämföra egenvärdena av deras korrelationskovariansmatriser, vilket erhåller en noggranhet på 84%. Att jämföra deras power delay profiles (PDP) råkar vara ett klassificeringsproblem av låg svårighetsgrad på grund av tidsupplösningsbegränsningar för de riktiga mätningarna. Vi vill påstå att den tilltänkta MATLAB- modellen aviker mycket från den givna datan. / Kandidatexjobb i elektroteknik 2020, KTH, Stockholm

In Wireless Sensor Actor Networks (WSANs), sensor nodes perform the sensing task and actor nodes take action based on the sensed phenomenon. The presence of actors in this configuration can not be benefited from, unless they are able to execute actions at right place and right time in the event region. The right place can be related to the accurate position of the sensor nodes. While, the right time is related to delivering the packets directly to the appropriate actors within the event specific response times. Hence, the efficient localization of sensor nodes, sensor-actor/actor-actor coordination and real-time routing is indispensable in WSANs. Furthermore, the limited energy levels and bandwidth of the state of art sensor nodes currently impose stringent requirements for low-complexity, low-energy, distributed coordination and cooperation protocols and their implementation. In this study, we propose an integrated framework which addresses the issues of sensors localization, network configuration, data aggregation, real-time data delivery, sensor-actor/actor-actor coordination and energy saving mechanisms. The proposal incorporates novel approaches on three fronts / (1) timing-based sensors localization (TSL) algorithm to localize the sensor nodes relative to actors, (2) real-time coordination and routing protocols and (3) energy conservation. The distributed real-time coordination and routing is implemented in addressing and greedy modes routing. A cluster-based real-time coordination and routing (RCR) protocol operates in addressing mode. The greedy mode routing approach (Routing by Adaptive Targeting, RAT) is a stateless shortest path routing. In dense deployment, it performs well in terms of delay and energy consumption as compared to RCR. To keep the traffic volume under control, the framework incorporates a novel real-time data aggregation (RDA) approach in RCR such that the packets deadlines are not affected. RDA is adaptive to the traffic conditions and provides fairness among the farther and nearer cluster-heads. Finally, framework incorporates a power management scheme that eliminates data redundancy by exploiting the spatial correlation of sensor nodes. Simulation results prove that the framework provides the real-time guarantees up to 95 % of the packets with lesser energy consumption of up to 33 % achieved using MEAC as compared to LEACH and SEP. The packet delivery ratio is also 60 % higher than that of semi-automated architecture. Furthermore the action accuracy is supported by TSL which restricts the localization errors less than 1 meter by tuning it according to the expected velocity of nodes and required accuracy.

Wireless multimedia sensor networks (WMSNs) are networks of interconnected devices that allow retrieving video and audio streams, still images, and scalar data from the environment. In a densely deployed WMSN, there exists correlation among the observations of camera sensors with overlapped coverage areas, which introduces substantial data redundancy in the network. In this dissertation, efficient communication schemes are designed for WMSNs by leveraging the correlation of visual information observed by camera sensors. First, a spatial correlation model is developed to estimate the correlation of visual information and the joint entropy of multiple correlated camera sensors. The compression performance of correlated visual information is then studied. An entropy-based divergence measure is proposed to predict the compression efficiency of performing joint coding on the images from correlated cameras. Based on the predicted compression efficiency, a clustered coding technique is proposed that maximizes the overall compression gain of the visual information gathered in WMSNs. The correlation of visual information is then utilized to design a network scheduling scheme to maximize the lifetime of WMSNs. Furthermore, as many WMSN applications require QoS support, a correlation-aware QoS routing algorithm is introduced that can efficiently deliver visual information under QoS constraints. Evaluation results show that, by utilizing the correlation of visual information in the communication process, the energy efficiency and networking performance of WMSNs could be improved significantly.

Image analysis is very important because of its usage in many different areas of science and industry. Pattern recognition (classification) is a tool used in image analysis. Statistical pattern recognition, based on Bayes discriminant functions is the object of this work. The main problem is to classify stationary Gaussian random field observation into one off two classes, considering, that it is dependant on training sample ant taking in to account the relationship with training sample. The new supervised classification method, based on Bayes discriminant functions, is proposed and it gives better results comparing with other commonly used Bayes discriminant functions. Method is programmed with R program and investigated experimentally, reconstructing images corrupted by spatially correlated noise. Such situation occurs naturally, for example, during the forest fire smoke covers the remotely sensed image, gathered from the satellite. Also such situation is often during cloudy days. During such situation the incorporation of the spatial dependences into the classification problem is useful. Analytical error rates of Bayes discriminant functions are presented (derived), which are the criterion of these functions. Also, the dependences on statistical parameters are investigated for these error rates. / Vaizdų analizė šiomis dienomis yra labai svarbi dėl plataus pritaikymo daugelyje mokslo ir pramonės sričių. Vienas iš vaizdų analizės įrankių – objekto atpažinimas (klasifikavimas) (angl. pattern recognition). Statistinis objekto atpažinimas, paremtas Bajeso diskriminantinėmis funkcijomis – šio darbo objektas. Sprendžiama problema – optimalus klasifikavimas stacionaraus Gauso atsitiktinio lauko (GRF) stebinio, į vieną iš dviejų klasių, laikant, kad jis yra priklausomas nuo mokymo imties ir atsižvelgiant į jo ryšius su mokymo imtimi. Pateikta klasifikavimo procedūra, kuri Gauso atsitiktinio lauko stebinius klasifikuoja optimaliai. Yra pasiūlytas naujas klasifikavimo su mokymu metodas, kuris duoda geresnius rezultatus, lyginant su įprastai naudojamomis Bajeso diskriminantinėmis funkcijomis. Metodas realizuotas R sistemos aplinkoje ir tikrinamas eksperimentų būdu, atstatant vaizdus, sugadintus erdvėje koreliuoto triukšmo. Tokia situacija pasitaiko natūraliai, pavyzdžiui, degant miškui dūmai uždengia nuotolinio stebėjimo vaizdą, gautą iš palydovo. Taip pat tokia situacija gana dažna esant debesuotumui. Esant tokiai situacijai erdvinės priklausomybės įvedimas į klasifikacijos problemą pasiteisina. Pateiktos (išvestos) analitinės klaidų tikimybių išraiškos Bajeso diskriminantinėms funkcijoms, kurios yra kaip šių funkcijų veikimo kriterijus. Ištirta klaidų tikimybių priklausomybė nuo statistinių parametrų reikšmių.

The class of models which is studied in this paper, multi-unit longitudinal models, combines both the cross-sectional and the longitudinal aspects of observations. Many empirical investigations involve the analysis of data structures that are both cross-sectional (observations are taken on several units at a specific time period or at a specific location) and longitudinal (observations on the same unit are taken over time or space). Multi-unit longitudinal data structures arise in economics and business where panels of subjects are studied over time, biostatistics where groups of patients on different treatments are observed over time, and in situations where data are taken over time and space. Modelling issues in multi-unit longitudinal models with random coefficients and patterned correlation structure are illustrated in the context of two data sets. The first data set deals with short time series data on annual death rates and alcohol consumption for twenty-five European countries. The second data set deals with glaceologic time series data on snow temperature at 14 different locations within a small glacier in the Austrian Alps. A practical model building approach, consisting of model specification, estimation, and diagnostic checking, is outlined. (author's abstract) / Series: Forschungsberichte / Institut für Statistik