Ph.D. Theses

Permanent URI for this collection

https://digitalarchive.library.bogazici.edu.tr/handle/123456789/11648

Browse

Now showing 1 - 20 of 95

A generalized pareto front approach for performance estimation in analog design automation systems
(Thesis (Ph.D.)-Bogazici University. Institute for Graduate Studies in Science and Engineering, 2010., 2010.) Deniz, Engin.; Dündar, Günhan,
The design of analog circuits requires a deep insight into both physical and technological parameter interactions. In addition, every application of analog circuits has speci c constraints and requirements with large number of design equations, and also there is no unique solution of the design. At the same time, analog circuits are the key components of mixed-signal systems. Nowadays, not only design issues but also tradeo analysis of competing performances is considered to be a signi cant issue in circuit modeling such that analog design automation tools, which increase e ciency and productivity, have become an attractive solution for integrated circuits (IC) providers. Furthermore, performance estimation tool becomes a requirement in order to speed up the automation system by eliminating the unfeasible circuits and the circuits which cannot meet the speci cations. In this thesis, a general methodology for the performance estimation of mixedsignal systems is proposed while exploiting the Pareto Front concept. Performance estimation requires a well-determined performance design space (PDS) exploration for a given technology. Since the complexity of mixed-signal systems grows progressively, the exploration of a huge design space is required for the performance estimation of system blocks with a dramatically increased exploration time. Therefore, a Matlabbased library is presented for a fast and accurate PDS exploration. Then, Pareto Front approach is applied to the system blocks. In addition, not only are optimum solution sets extracted but also an approximate design of the blocks is obtained in this thesis. Finally, Pareto Front composition is discussed by supporting the dominance rule with algebraic representations.
A high voltage triboelectric energy harvesting system utilizing parallel-SSHI rectifier and DC-DC converters for SUB-5 hz motions
(Thesis (Ph.D.) - Bogazici University. Institute for Graduate Studies in Science and Engineering, 2020., 2020.) Kara, İsmail.; Mutlu, Şenol.; Doğan, Hakan.
In this thesis, the first integrated circuit (IC) implementation of parallel synchronized switching harvesting on inductor (parallel-SSHI) is presented for triboelectric energy harvester targeting 1 Hz to 5 Hz mechanical motions. It is accompanied by on-chip buck and switched-capacitor DC-DC converters, all capable of handling 70 V levels. Unlike piezoelectric harvesters, triboelectric nanogenerators (TENGs) can produce very high open-circuit voltages; thus, the proposed system utilizes this property within the technology limits to maximize the extracted power. An in-house manufactured TENG using steel and polytetra uoroethylene (PTFE) is modeled for sub-5 Hz motions. The energy is extracted and stored in an external capacitance until its voltage reaches 70 V, which is achieved in three press-and-release mechanical cycles. 70-to-2 V down conversion is carried on by a 70-to-10 V buck converter followed by a 10-to-2 V switched-capacitor DC-DC converter. A chip is manufactured in TSMC 0.18 m HV BCD process with an active area of 6.25 mm2. End-to-end peak e ciency is measured as 32.71% for 1 Hz motion with a 722 W total power delivery to the load for 4 ms.
Achievable rates and transceiver design in ultra-wideband communications
(Thesis (Ph.D.)-Bogazici University. Institute for Graduate Studies in Science and Engineering, 2009., 2009.) Güney, Nazlı.; Deliç, Hakan.
In a multipath dominated environment, ultra-wideband (UWB) systems that transmit trains of subnanosecond duration pulses exhibit the desirable property of fine resolution in time of the received paths, which as a result of the impulsive form of the transmitted signal go through fewer amplitude fluctuations than those emanating from systems with narrower bandwidths. Being distributed over a large number of resolvable paths, UWB signal energy is typically collected by the rake receiver. In this thesis, achievable information rates of time-hopping M-ary pulse position modulation UWB systems using either soft- or hard-decision outputs are calculated first, where one distinguishing characteristic observed for the hard-output systems is that increasing the constellation size is advantageous only at sufficiently large values of the code rate. Next, it is shown that with time division duplex UWB systems, for which channel information is available at the transmitter, it is possible to move about half of the rake fingers to the transmitter, and simultaneously increase the received signal-to-noise ratio (SNR). The impact of the nature of the noise phenomenon on the rake receiver is such that clipper nonlinearities following the rake fingers are needed if non-Gaussian noise is present. To this end, a robust rake receiver is designed and its performance is optimized through the parameters of the nonlinearities. Finally, a robust multipath-combining decorrelating (mD) detector is developed for non-Gaussian channels. Corresponding to a structure with further processing of the rake receiver outputs, the robust mD detector effectively removes the interference from the other users as well as the impulsive noise, and thus the error floor observed with rake receivers and single-user detection at high SNR values and large number of users is avoided.
Adaptive control of nonlinear systems using multiple identification models
(Thesis (Ph.D.)-Bogazici University. Institute for Graduate Studies in Science and Engineering, 2007., 2007.) Cezayirli, Ahmet.; Cılız, Kemal.
Adaptive control of nonlinear systems is considered in this study. Available methods in this field are reviewed first. Focusing on the minimum-phase, input-output linearizable and linearly parameterized nonlinear systems, direct and indirect adaptive controllers are developed for the cases of matched and unmatched uncertainties. A new methodology is proposed, which makes use of multiple identification models in order to improve the transient performance under large parametric uncertainties. Adaptation and switching mechanisms are developed based on the use of multiple Lyapunov functions and cost functions. The resulting closed loop systems are shown to be stable with the switching mechanisms. Combination of direct and indirect adaptive control schemes is also presented for a class of nonlinear systems which do not give rise to over-parametrization. The theoretical results obtained in this study are verified by computer simulations.
Advanced techniques for the design of MIMO free space optical communication systems
(Thesis (Ph.D.) - Bogazici University. Institute for Graduate Studies in Science and Engineering, 2016., 2016.) Özbilgin, Tuğba.; Koca, Mutlu.
Free space optical (FSO) communication is an effective alternative to radio frequency (RF) communication in terms of potential high data rates, cost efficiency and wide bandwidth on unregulated spectrum. However, FSO systems are susceptible to severe atmospheric turbulence conditions causing random fluctuations in the intensity and phase of the optical signal. These turbulence induced fading effects may severely degrade the FSO system performance, limiting its use to only short-range applications and clear weather conditions. In this dissertaion, with the motivation to overcome these limitations, analysis and design of FSO communication systems are considered. First a unified framework for determining the inter-aperture separations in multiple-input-multiple-output (MIMO) FSO systems such that the transmitterreceiver paths are resolvable is presented. Second a novel non-coherent MIMO optical modulation method combining optical spatial modulation (OSM) with joint pulse position and amplitude modulations, which offers better power and spectral efficiencies than conventional MIMO optical modulation methods, is proposed. Then two novel FSO communication approaches based on polarization shift keying (POLSK) are considered. The first non-coherent approach combines MIMO POLSK with pulse position modulation (PPM) constituting a good compromise between power and spectral efficiencies compared to standard uses of PPM and POLSK, while the second employs POLSK in an OSM based MIMO architecture and introduces a spectrally efficient coherent FSO communication scheme that also offers better error performance gains than other spectrally efficient modulation techniques. Finally error ratios belonging to separate signalling dimensions of these POLSK schemes are evaluated and unequal error protection (UEP) is offered to optimize the error control coding gains attained.
Advanced transceiver design for continuous phase modulation
(Thesis (Ph.D.)-Bogazici University. Institute for Graduate Studies in Science and Engineering, 2008., 2008.) Özgül, Barış.; Deliç, Hakan.; Koca, Mutlu.
This dissertation proposes advanced transceiver designs applying turbo and space-time (ST) concepts to continuous phase modulation (CPM), which is preferred in numerous power- and band-limited communication systems for its constant envelope and spectral e ciency. Despite its highly attractive spectral properties, maximumlikelihood detection of CPM over the frequency-selective multipath fading channels can bring impractical complexity issues because of the intensive search over a single super trellis which combines the e ects of the modulation and the multipath channel. Application of the reduced-state trellis search algorithms results in lower complexity but the computational load could still be prohibitively large to obtain high performance in long channel impulse responses. In the dissertation, instead of employing trellis-based combined detection methods, equalization and demodulation functions are separated and novel low-complexity receivers with soft-input soft-output (SISO) time-domain and frequency-domain linear equalizers are proposed for bit-interleaved coded CPM, which attain near-optimal performance by applying turbo processing. In the proposed receivers, the front-end soft-information-aided linear equalizer is followed by a central SISO CPM demodulator and a back-end SISO channel decoder where double turbo processing is employed by performing back-end demodulation/decoding iterations per each equalization iteration to improve the a priori information for the front-end equalizer. Performance for the frequency-domain equalization is further improved by proposing an orthogonal ST block coding scheme for CPM. The proposed technique maintains the constant envelope and the phase continuity of the CPM waveforms perfectly by using appropriate tail symbols and, therefore, has no impact on the spectral e ciency. Depending on the orthogonality of the ST combining, frequencydomain equalization is applied as in the case of single antenna transmissions without v any increase in the computational load. In the dissertation, the receiver complexity is reduced further by transferring all the equalization functions to the transmitter and employing pre-equalization. For precoding the CPM signals on multipath fading channels while maintaining the spectral e ciency, a novel ST pre-equalizer is proposed, limiting the envelope variations and attaining a peak-to-average power ratio that is close to one by using a transmit selection diversity scheme.
Alternative arithmetic structures using redundant numbers and multi-valued circuit techniques
(Thesis (Ph.D.)-Bogazici University. Institute for Graduate Studies in Science and Engineering, 2010., 2010.) Çini, Uğur.; Morgül, Avni,
Arithmetic circuits play a crucial role in VLSI technology. Arithmetic blocks are usually the most power consuming parts in a system since the switching activity is quite high. Alternative arithmetic implementations can be a solution to reduce power consumption and to increase the performance of the whole system. Static CMOS digital design has robust working performance, where logic levels are kept at the two extremes, either the ground voltage or supply voltage. However, the voltage excursion between the supply voltage and ground at all nodes causes excessive power dissipation. This condition also generates noise over the whole circuitry, which is not desirable especially in mixed signal designs. Current-mode digital design techniques can be a solution for this issue especially whenever the switching activity is high. In the first part of the thesis, alternative current-mode arithmetic structures are built focusing on multi-valued circuits. Together with multi-valued logic implementations, signed-digit numbers and redundant number systems are also analyzed. The design issues of multi-valued circuits are discussed and novel building blocks for multi-operand addition are developed. In the second part of the thesis, redundant arithmetic schemes for new generation reconfigurable systems are also analyzed. These techniques proposed here can be implemented efficiently by using recently introduced 6-input look-up table based field programmable gate array (FPGA) systems. A redundant double carry-save mode addition technique is proposed for the new generation FPGA devices. Using the proposed technique, efficient multiply-accumulate operations and finite impulse response filter structures for reconfigurable systems are developed.
Analog circuit design automation against process variations and aging phenomena
(Thesis (Ph.D.) - Bogazici University. Institute for Graduate Studies in Science and Engineering, 2016., 2016.) Afacan, Engin.; Başkaya, Faik.
Reliability of CMOS circuits has become a major concern due to substantially worsening process variations and aging phenomena in deep sub-micron devices. As a result, conventional analog circuit sizing tools have become incapable of promising a certain yield whether it is immediately after production or after a certain period of time. Thereby, analog circuit sizing tools have been replaced by better ones, where reliability is included in the conventional optimization problem. Variation-aware analog circuit synthesis has been studied for many years, and numerous methodologies have been proposed in the literature. On the other hand, as far as we know, there has not been any tool that takes lifetime into account during the optimization. Besides, there are a number of di erent issues with lifetime-aware circuit optimization, where aging analysis is still quite problematic due to modeling and simulation de ciencies. Furthermore, both tools su er from the challenging trade-o between e ciency and accuracy. Recon gurable analog circuit design is another way of designing analog circuits against aging. However, design of a such complicated system is highly time consuming process to be performed by hand. Even though recon gurable circuit design has been studied in the literature, there has been no attempt to automatize the design process to reduce the design time. With regard to aforementioned these problems, this study addresses all of these problems under a general title of reliability-aware analog circuit design automation, severally discusses them in detail, and proposes novel solutions to deal with not only existing but also not addressed problems.
Asynchronous sigma-delta analog to digital converters
(Thesis (Ph.D.)-Bogazici University. Institute for Graduate Studies in Science and Engineering, 2015., 2015.) Kayaaltı, Balkır.; Cerid, Ömer.
Analog to digital conversion is the process of translating analog real world signals to their corresponding binary encoded form which digital devices need. Among the examples for analog signal sources storage devices like harddisks and sensor devices can be given. Analog to digital converters are key components that make this translation. A specific topology can be selected according to its achievable number of bits, speed, power and according to its area of usage. Sigma-delta analog to digital conversion is a popular type which can be used in applications where high resolution with moderate bandwidth is desired. High resolutions are achieved by the inherent noise suppression by the loop filter. Oversampling, which corresponds to the sampling of the input signal with a higher rate than that of the regular Nyquist rate is their other advantage. This process pushes the quantization noise away from the baseband, and thereby higher signal to noise ratios are achieved. There are mainly two sub-types of sigma delta converters. The first, is the discrete time version which uses discrete time signals and components. The second is the continuous time version which uses entire continuous time signals and components except the quantizer and the digital to analog converter (DAC) in the feedback. These two types are both clocked with an external clock signal, so the output is synchronous i.e. synchronized with a clock. Designing a new type of the circuit which does not use an external clock; but instead using the self oscillation created in the loop can be a an alternative third type. This type of the converter will not use sampling inside the loop, so clock jitter due to the sampling can be eliminated , where usually a high frequency signal is sampled. If the sampling is done with a lower frequency, effect of the clock jitter can be reduced. This thesis is about the theory of such converters and their design procedure with a real circuit implementation.
Autonomus multi-robot topological spatial cognition
(Thesis (Ph.D.) - Bogazici University. Institute for Graduate Studies in Science and Engineering, 2015., 2015.) Karaoğuz, Hakan.; Bozma, H. Işıl.
This thesis is concerned with topological spatial cognition in multi-robot systems. With topological reasoning, continuous space is discretized into a set of places that are spatially related. Thus, topological spatial cognition is associated with the acquisition, organization, utilization and revision of knowledge of places and their spatial relations. In turn, this can occur either through the direct experience of the robot or indirectly based on the knowledge of other robots. In this perspective, the problem is handled in three stages. First, efficient sensory data representation is studied and a model based on previously developed bubble space is presented. Next, the full range of spatial processing associated with direct experience is considered via introducing a topological spatial cognition (TSC) model. This model enables each robot to detect places, recognize them or learn them as necessary in a completely unsupervised, incremental and organized manner. The robot continually builds and utilizes its long-term spatial memory where knowledge of places and their spatial relations are retained in separate, but related parts. Finally, the problem of expanding each robot’s spatial cognition based on other robots’ knowledge is addressed and a model that enables each robot to merge its spatial memory with those of other robots is proposed. All of the proposed approaches are evaluated on benchmark data sets as well as on real robots. The experimental results demonstrate that robots are able to autonomously become cognizant of their surrounding through either their individual experience or that of other robots.
Broadband power-line noise modeling and compensation
(Thesis (Ph.D.)-Bogazici University. Institute for Graduate Studies in Science and Engineering, 2011., 2011.) Güzel, Tayyar.; Mıhçak, Mehmet Kıvanç.; Deliç, Hakan.
Broadband power-line communication (BPLC) has become more popular in the Smart Grid context as it allows broadband access, media sharing, automated meter reading etc. However, it still remains as a challenging task to characterize and compensate for the impairments of the power-line networks for high-speed communications. In the scope of the thesis study, multi-path fading and path loss of the powerline channel are studied and a unit length attenuation profile (ULAP) is derived from measurements as part of the channel characterization. Using ULAP and the other channel parameters, a power-line channel simulator is developed. Correctness of the simulation is verified by measurement. Using the PLC channel simulations, statistics of the channel parameters are also studied. During the course of the thesis work, both broadband noise and home appliance noise are studied and characterized. Background noise is analyzed in frequency domain and Generalized Gaussian distribution (GGD) is proposed for modeling the noise amplitude distribution of each individual frequency. The proposed model is verified by measurements in the Kullback-Leibler sense of distribution mismatch penalty. Using the proposed noise model, a novel Orthogonal Frequency Division Multiplexing (OFDM) based receiver is proposed. It is shown that the performance of the proposed receiver utilizing the parameters of the proposed noise model reveals promising results.
Channel polarization with higher - order memory
(Thesis (Ph.D.) - Bogazici University. Institute for Graduate Studies in Science and Engineering, 2016., 2016.) Afşer, Hüseyin.; Deliç, Hakan.
Channel coding is one of the most fundamental problems regarding the nature of communication. Polar codes, developed by Ar kan, were the rst demonstration of practical channel codes that provably achieve theoretical limits in a wide range of communication scenarios. Based on simple channel transformations which are called channel combining and splitting, Ar kan developed the idea of channel polarization which resulted in polar codes. In this dissertation we focus on the unique properties of polar codes and channel polarization. First, we investigate the channel-speci c construction of polar codes, a point which discriminates polar codes from Reed-Muller (RM) codes. We obtain results showing the inherent e ect of the underlying channel on the construction of polar codes and provide a bridge between polar and RM codes. Our results easily extend to obtaining a characterization for the rate of polarization as well. Next, we consider practical uses of polar codes for fading channels. We design a bit-interleaved polar-coded modulation scheme (BIPCM) by deriving a low complexity code-construction method and designing a lower complexity successive cancellation list decoder (SCLD). We compare the resultant BIPCM system with the existing solutions and show that it provides signi cant performance advantages. Finally, we generalize the channel polarization idea by changing channel combining and splitting operations. By introducing a memory order in the channel combining process we obtain a class of codes, including the original ones, that are parametrized by the memory order. We show that the new family of polar codes achieve the theoretical limits as well and they can also be used with lower complexity by increasing the memory order. We thereby complement Ar kan's conjecture that channel polarization is in fact a general phenomenon.
Circuit level analog design automation
(Thesis (Ph.D.)-Bogazici University. Institute for Graduate Studies in Science and Engineering, 2010., 2010.) Sönmez, Özsun Serkan.; Dündar, Günhan,
This thesis presents a simulation-based analog circuit synthesis methodology, its integration with system and layout level analog synthesis tools, and synthesis examples that were performed to validate the usefulness of the methodology. Simulation-based approach is preferred so that the synthesizer, SACSES, is topology independent. Instead of using a commercially available simulator, an accelerated simulator, SPASE, is implemented. SPASE has various acceleration mechanisms for DC, AC and noise simulation, detailed in the thesis. The search algorithm used is an ES algorithm modified so as to use Metropolis criterion as the selection method. By adding hill-climbing capability similar to simulated annealing, this modification provides a proper balance between local and global search and eliminates the premature convergence problem of GA and ES-based algorithms. Designer effort is minimized with the automated determination and self-evolution of search parameters. Smooth penalty mechanisms for biasing constraints are proposed and embedded in the algorithm. Yield-aware synthesis is performed by utilizing piecewise cubic Hermite splines for response surface modeling. A hierarchical synthesis structure is proposed for integrating SACSES with system level synthesis tools. The hierarchical scheme eliminates the need for extra tools to link levels. SACSES and the layout level synthesis tool is integrated with a feedback loop, by which the e®ects of layout parasitics on circuit performance are minimized.
CMOS based optoelectronic catheter localization system for magnetic resonance imaging environment
(Thesis (Ph.D.) - Bogazici University. Institute for Graduate Studies in Science and Engineering, 2013., 2013.) Sarıoğlu, Baykal.; Yalçınkaya, Arda Deniz.
The use of minimally invasive medical techniques has greatly reduced risks to patients and has increased our understanding of how the body works. Detection of catheters is one of the most important tasks in these techniques. Magnetic resonance imaging provides a safe environment for catheter driven operations. On the other hand, implementation of catheter localization architectures are problematic due to the impossibility of long conductor usage due to RF induced heating problem in MRI environment. There is a need for devices that use transmission systems other than electrical for realizing catheter driven operations in MRI environment. This thesis presents design and implementation of a micro-system that uses optical transmission for localization of catheters in MRI environment. The micro-system is composed of a radio frequency (RF) integrated circuit (IC) designed in 0.18 m triple well CMOS technology and optoelectronic components. The RF integrated circuit covers an RF receiver architecture and an optical power supply unit that provides power to the receiver. The power to the system is served by a laser with a wavelength of 660 nm. RF receiver is connected to a micro-coil antenna with a diameter of 2 mm, and it can transmit very low powered magnetic resonance signals to the external environment via a ber-coupled LED with a wavelength of 1310 nm. The external optical receiver and signal processing unit detect the location of the catheter using the frequency information of the received signal. The implemented system is tested successfully in 3T MRI environment. The tests show that the system has 350 m spatial resolution, which proves that the implemented system can be used for the catheter driven minimally invasive operations.
Consensus based power control algorithms for heterogenous networks
(Thesis (Ph.D.) - Bogazici University. Institute for Graduate Studies in Science and Engineering, 2016., 2016.) Şenel, Kamil.; Akar, Mehmet.
The resource allocation problem is crucial for unveiling the potential of heterogeneous networks. The demand for higher data rates from an increasing number of devices compels the operators to look for solutions beyond the traditional network architecture. Heterogeneous network architecture is a promising solution to provide data rates required by the emerging applications and high-speed multimedia services. In this dissertation, we focus on the interference management for heterogeneous networks which is a major problem to be resolved and propose solutions based on power control techniques. We utilize the inherent advantages of consensus algorithms such as non-essentially of objective functions and fairness to design power control algorithms suitable for the task of resource allocation in heterogeneous networks. In this dissertation, several novel, distributed and self optimized power adjustment algorithms are proposed. Contrary to the approaches in the literature, the instantaneous or statistical measurements on channel gains are not required during the power adjustment process. The convergence analyses reveal that the proposed algorithms achieve optimum solution for the power allocation problem with fairness constraints, even under a setup with imperfect communication links. Furthermore, the theoretical analyses show that the convergence properties of the proposed algorithms are preserved under di erent spectrum allocation schemes. The numerical analyses are in agreement with the theoretical analyses and demonstrate signi cant improvement in terms of overall network performance.
Content based medical image retrieval
(Thesis (Ph.D.) - Bogazici University. Institute for Graduate Studies in Science and Engineering, 2017., 2017.) Marvasti, Neda Barzegar.; Acar, Burak.
Fast technological developments of di erent medical imaging and data collection techniques increase the expectation of more accurate interpretations and diagnoses of radiologists. However, to carefully analyze the resulting big medical data, reliable and fast systems are needed. Content-based medical image retrieval (CBMIR) is a valuable technique to assist radiologists by identifying similar images in a large archive. However, due to the huge semantic gap between low-level image features and high-level semantic features, the challenge of retrieving similar images utilizing the high-level user speci ed semantic labels, which are closer to the users understandings and interpretations, has attracted great interest from various researches. In this dissertation, an iterative search and retrieval scheme to identify similar images from a database of 3-dimensional liver computed tomography (CT) images is proposed via utilizing the combination of lesion and liver related semantic features and patients' metadata. At each retrieval iteration, the lesion related concepts are annotated in a speci c order through a proposed computer aided medical image annotation (CMIA) scheme. The proposed radiologist-in-the-loop semi-automatic CMIA is based on a Bayesian tree structured model, linked to RadLex, to exploit the inter-dependencies between concepts to update the full annotation process and to guide the radiologist to input the most critical information at each iteration. Results show the e ectiveness of this modelbased interactive annotation scheme compared to the domain-blind models, as well as its advantage in the performance of the retrieval system, where a few number of manual annotations can signi cantly boost the retrieval accuracy. Moreover, better retrieval performance is achieved by incorporating a small contribution of the non-lesion data.
Convergence rate analysis and optimization of distributed consensus algorithms
(Thesis (Ph.D.) - Bogazici University. Institute for Graduate Studies in Science and Engineering, 2014., 2014.) Cihan, Onur.; Akar, Mehmet.
The problem of achieving a common value in a distributed information sharing context, referred to as distributed consensus or agreement, is an important topic that has drawn significant research attention of late. Consensus algorithms find applications in many areas including network clock synchronization, sensor fusion and load balancing where achieving consensus as fast as possible is important. In this dissertation, we study the analysis and optimization of convergence rate of averaging based distributed consensus algorithms evolving on graphs. By relating the convergence speed of the algorithm to the mixing rate of a Markov chain, we propose two semi–definite programming (SDP) methods of assigning transition probabilities to a Markov chain in order to optimize its mixing rate. In the first SDP formulation, there is a single transition probability parameter to be optimized (the holding probability of vertices) which leads to easier and faster computation as opposed to the more general reversible Markov chain formulation corresponding to a stationary distribution that is proportional to the degree of vertices. By deriving exact analytical results, it is shown that both the single parameter and the degree proportional reversible fastest mixing Markov chain formulations yield better results than the symmetric SDP formulation for a path and some well–known edge–transitive and orbit graphs. The convergence rate of the averaging based distributed consensus algorithm is also analyzed for networks where delay exists in data receptions, which is unavoidable in practice. After introducing the delayed version of the consensus algorithm, it is analytically shown that bounded non–uniform delay does not adversely affect its convergence rate for directed acyclic graphs.
Cross-lingual voice conversion
(Thesis (Ph.D.)-Bogazici University. Institute for Graduate Studies in Science and Engineering, 2007., 2007.) Türk, Oytun.; Arslan, Levent M.
Cross-lingual voice conversion refers to the automatic transformation of a source speaker’s voice to a target speaker’s voice in a language that the target speaker can not speak. It involves a set of statistical analysis, pattern recognition, machine learning, and signal processing techniques. This study focuses on the problems related to cross-lingual voice conversion by discussing open research questions, presenting new methods, and performing comparisons with the state-of-the-art techniques. In the training stage, a Phonetic Hidden Markov Model based automatic segmentation and alignment method is developed for cross-lingual applications which support textindependent and text-dependent modes. Vocal tract transformation function is estimated using weighted speech frame mapping in more detail. Adjusting the weights, similarity to target voice and output quality can be balanced depending on the requirements of the cross- lingual voice conversion application. A context-matching algorithm is developed to reduce the one-to-many mapping problems and enable nonparallel training. Another set of improvements are proposed for prosody transformation including stylistic modeling and transformation of pitch and the speaking rate. A high quality cross-lingual voice conversion database is designed for the evaluation of the proposed methods. The database consists of recordings from bilingual speakers of American English and Turkish. It is employed in objective and subjective evaluations, and in case studies for testing new ideas in cross- lingual voice conversion.
Current-model CMOS sequential multiple-valued logic circuits
(Thesis (Ph.D.)-Bogazici University. Institute for Graduate Studies in Science and Engineering, 2012., 2012.) Sarıca, Fatma.; Morgül, Avni,
The use of circuits with more than two logic levels, named as Multiple-Valued Logic (MVL) circuits, have a potential for reducing chip area consumed by interconnection wiring and functional units in Very-Large Scale Integration (VLSI). Many logical and arithmetic functions have been shown to be more efficiently implemented with multiple-valued logic in terms of number of operations, gates, transistor count and signal lines etc. However, in spite of their potential advantages, developments in multi-valued systems are not satisfactory. It is still a complicated task to design a system for processing a signal in a multi-valued manner despite considerable effort. Multi-valued logic circuits can be designed in current mode or voltage mode. Due to the limited supply voltages, higher radices could not be obtained using voltage mode circuits. On the other hand, current mode circuits have the advantages of current scaling, copying and sign changing with a simple current mirror, but unlike binary logic circuits, they are not self-restored. In this study, current-mode sequential logic circuits are designed and analyzed. Multiple-valued counterparts of the well-known flip-flop structures are discussed and a new type of flip-flop circuit, named AB, is proposed. The proposed circuit is used in various counter and random sequential circuit designs. Circuit schematics and simulation results are presented. In addition, a multiple-valued and binary counter designs that are using D-type flip-flops (the only flip-flop that has the same output equations for both MVL and binary) are compared in terms of various VLSI design criteria.
DDoS attack detection using signal processing and statistical approaches
(Thesis (Ph.D.) - Bogazici University. Institute for Graduate Studies in Science and Engineering, 2021., 2021.) Erhan, Derya.; Anarım, Emin.
DDoS attacks cause a variety of changes in the properties of the attributes in the network traffic. Modeling these changes using signal processing and statistical ap proaches provides detection of these attacks. This thesis focuses on detecting DDoS attacks using time series analysis, sparse signal representation methods, and statistical modeling. We also investigate the effect of DDoS attacks on traffic features in a sta tistical manner. In addition, we propose two simple but effective network-based DDoS attack detection methods based on the statistical signal processing approach, using the advantage of statistical changes in traffic features. We propose a novel DDoS detection framework using the Matching Pursuit algo rithm to detect resource depletion type DDoS attacks. We use multiple characteristics of network traffic simultaneously to detect low-density DDoS attacks efficiently. The proposed method uses the dictionary produced from the parameters of the network traf fic using the K-SVD algorithm. Dictionary generation using network traffic provides legitimate and attack traffic models and adds adaptability to the proposed method to network traffic. We also implement DDoS detection approaches that use Matching Pursuit and Wavelet techniques and compare them using two different data sets. Addi tionally, we offer a hybrid DDoS detection framework that combines these approaches with a decision-making mechanism using an artificial neural network. We evaluate the proposed methods with two different data sets. In the hybrid intrusion detection sys tem with more than one attack, the detection performances of other approaches have decreased. In contrast, the proposed method achieves true-positive rates higher than 99% with a false positive rate lower than 0.7%.

Browse

Browsing Ph.D. Theses by Title

Results Per Page

Sort Options