Electrical - Electronic Engineering / Elektrik - Elektronik Mühendisliği
Permanent URI for this collectionhttps://hdl.handle.net/11147/11
Browse
9 results
Search Results
Now showing 1 - 9 of 9
Article Citation - WoS: 22Citation - Scopus: 26Adaptive Sign Algorithm for Graph Signal Processing(Elsevier, 2022) Yan, Yi; Kuruoğlu, Ercan Engin; Altınkaya, Mustafa AzizEfficient and robust online processing techniques for irregularly structured data are crucial in the current era of data abundance. In this paper, we propose a graph/network version of the classical adaptive Sign algorithm for online graph signal estimation under impulsive noise. The recently introduced graph adaptive least mean squares algorithm is unstable under non-Gaussian impulsive noise and has high computational complexity. The Graph-Sign algorithm proposed in this work is based on the minimum dispersion criterion and therefore impulsive noise does not hinder its estimation quality. Unlike the recently proposed graph adaptive least mean pth power algorithm, our Graph-Sign algorithm can operate without prior knowledge of the noise distribution. The proposed Graph-Sign algorithm has a faster run time because of its low computational complexity compared to the existing adaptive graph signal processing algorithms. Experimenting on steady-state and time-varying graph signals estimation utilizing spectral properties of bandlimitedness and sampling, the Graph-Sign algorithm demonstrates fast, stable, and robust graph signal estimation performance under impulsive noise modeled by alpha stable, Cauchy, Student's t, or Laplace distributions.Article Citation - WoS: 1Maximum Average Entropy-Based Quantization of Local Observations for Distributed Detection(Elsevier, 2022) Wahdan, Muath A.; Altınkaya, Mustafa AzizIn a wireless sensor network, multilevel quantization is necessary to find a compromise between minimizing the power consumption of sensors and maximizing the detection performance at the fusion center (FC). The previous methods have been using distance measures such as J-divergence and Bhattacharyya distance in this quantization. This work proposes a different approach based on the maximum average entropy of the output of the sensors under both hypotheses and utilizes it in a Neyman-Pearson criterion-based distributed detection scheme to detect a point source. The receiver operating characteristics of the proposed maximum average entropy (MAE) method in quantizing sensor outputs have been evaluated for multilevel quantization both when the sensor outputs are available error-free at the FC and when non-coherent M-ary frequency shift keying communication is used for transmitting MAE based multilevel quantized sensor outputs over a Rayleigh fading channel. The simulation studies show the success of the MAE in the cases of both error-free fusion and where the effect of the wireless channel has been incorporated. As expected, the performance improves as the level of quantization increases and with six-level quantization approaches the performance of non-quantized data transmission.Article Citation - WoS: 4Citation - Scopus: 5Beyond Trans-Dimensional Rjmcmc With a Case Study in Impulsive Data Modeling(Elsevier Ltd., 2018) Karakuş, Oktay; Kuruoğlu, Ercan Engin; Altınkaya, Mustafa AzizReversible jump Markov chain Monte Carlo (RJMCMC) is a Bayesian model estimation method, which has been generally used for trans-dimensional sampling and model order selection studies in the literature. In this study, we draw attention to unexplored potentials of RJMCMC beyond trans-dimensional sampling. the proposed usage, which we call trans-space RJMCMC exploits the original formulation to explore spaces of different classes or structures. This provides flexibility in using different types of candidate classes in the combined model space such as spaces of linear and nonlinear models or of various distribution families. As an application, we looked into a special case of trans-space sampling, namely trans-distributional RJMCMC in impulsive data modeling. In many areas such as seismology, radar, image, using Gaussian models is a common practice due to analytical ease. However, many noise processes do not follow a Gaussian character and generally exhibit events too impulsive to be successfully described by the Gaussian model. We test the proposed usage of RJMCMC to choose between various impulsive distribution families to model both synthetically generated noise processes and real-life measurements on power line communications impulsive noises and 2-D discrete wavelet transform coefficients.Article Citation - WoS: 103Citation - Scopus: 122One-Day Ahead Wind Speed/Power Prediction Based on Polynomial Autoregressive Model(Institution of Engineering and Technology, 2017) Karakuş, Oktay; Kuruoğlu, Ercan Engin; Altınkaya, Mustafa AzizWind has been one of the popular renewable energy generation methods in the last decades. Foreknowledge of power to be generated from wind is crucial especially for planning and storing the power. It is evident in various experimental data that wind speed time series has non-linear characteristics. It has been reported in the literature that nonlinear prediction methods such as artificial neural network (ANN) and adaptive neuro fuzzy inference system (ANFIS) perform better than linear autoregressive (AR) and AR moving average models. Polynomial AR (PAR) models, despite being non-linear, are simpler to implement when compared with other non-linear AR models due to their linear-in-the-parameters property. In this study, a PAR model is used for one-day ahead wind speed prediction by using the past hourly average wind speed measurements of Ceşme and Bandon and performance comparison studies between PAR and ANN-ANFIS models are performed. In addition, wind power data which was published for Global Energy Forecasting Competition 2012 has been used to make power predictions. Despite having lower number of model parameters, PAR models outperform all other models for both of the locations in speed predictions as well as in power predictions when the prediction horizon is longer than 12 h.Article Citation - WoS: 11Citation - Scopus: 10Bayesian Volterra System Identification Using Reversible Jump Mcmc Algorithm(Elsevier Ltd., 2017) Karakuş, Oktay; Kuruoğlu, Ercan Engin; Altınkaya, Mustafa AzizVolterra systems have had significant success in modelling nonlinear systems in various real-world applications. However, it is generally assumed that the nonlinearity degree of the system is known beforehand. In this paper, we contribute to the literature on Volterra system identification (VSI) with a numerical Bayesian approach which identifies model coefficients and the nonlinearity degree concurrently. Although this numerical Bayesian method, namely reversible jump Markov chain Monte Carlo (RJMCMC) algorithm has been used with success in various model selection problems, our use is in a novel context in the sense that both memory size and nonlinearity degree are estimated. The aforementioned study ensures an anomalous approach to RJMCMC and provides a new understanding on its flexible use which enables trans-structural transitions between different classes of models in addition to transdimensional transitions for which it is classically used. We study the performance of the method on synthetically generated data including OFDM communications over a nonlinear channel.Article Citation - WoS: 68Citation - Scopus: 74A Novel Acoustic Indoor Localization System Employing Cdma(Elsevier Ltd., 2012) Sertatıl, Cem; Altınkaya, Mustafa Aziz; Raoof, KosaiNowadays outdoor location systems have been used extensively in all fields of human life from military applications to daily life. However, these systems cannot operate in indoor applications. Hence, this paper considers a novel indoor location system that aims to locate an object within an accuracy of about 2 cm using ordinary and inexpensive off-the-shelf devices and that was designed and tested in an office room to evaluate its performance. In order to compute the distance between the transducers (speakers) and object to be localized (microphone), time-of-arrival measurements of acoustic signals consisting of Binary Phase Shift Keying modulated Gold sequences are performed. This DS-CDMA scheme assures accurate distance measurements and provides immunity to noise and interference. Two methods have been proposed for location estimation. The first method takes the average of four location estimates obtained by trilateration technique. In the second method, only a single robust position estimate is obtained using three distances while the least reliable fourth distance measurement is not taken into account. The system's performance is evaluated at positions from two height levels using system parameters determined by preliminary experiments. The precision distributions in the work area and the precision versus accuracy plots depict the system performance. The proposed system provides location estimates of better than 2 cm accuracy with 99% precision.Article Citation - WoS: 1Citation - Scopus: 1Benefits of Averaging Lateration Estimates Obtained Using Overlapped Subgroups of Sensor Data(Elsevier Ltd., 2013) Altınkaya, Mustafa AzizIn this paper, we suggest averaging lateration estimates obtained using overlapped subgroups of distance measurements as opposed to obtaining a single lateration estimate from all of the measurements directly if a redundant number of measurements are available. Least squares based closed form equations are used in the lateration. In the case of Gaussian measurement noise the performances are similar in general and for some subgroup sizes marginal gains are attained. Averaging laterations method becomes especially beneficial if the lateration estimates are classified as useful or not in the presence of outlier measurements whose distributions are modeled by a mixture of Gaussians (MOG) pdf. A new modified trimmed mean robust averager helps to regain the performance loss caused by the outliers. If the measurement noise is Gaussian, large subgroup sizes are preferable. On the contrary, in robust averaging small subgroup sizes are more effective for eliminating measurements highly contaminated with MOG noise. The effect of high-variance noise was almost totally eliminated when robust averaging of estimates is applied to QR decomposition based location estimator. The performance of this estimator is just 1 cm worse in root mean square error compared to the Cramér–Rao lower bound (CRLB) on the variance both for Gaussian and MOG noise cases. Theoretical CRLBs in the case of MOG noise are derived both for time of arrival and time difference of arrival measurement data.Article Citation - WoS: 4Citation - Scopus: 4Phase Dependence Mitigation for Autocorrelation-Based Frequency Estimation(Elsevier Ltd., 2008) Altınkaya, Mustafa Aziz; Anarım, Emin; Sankur, BülentThe sinusoidal frequency estimation from short data records based on Toeplitz autocorrelation (AC) matrix estimates suffer from the dependence on the initial phases of the sinusoid(s). This effect becomes prominent when the impact of additive noise vanishes, that is at high signal-to-noise ratios (SNR). Based on both analytic derivation of the AC lag terms and simulation experiments we show that data windowing can mitigate the limitations caused by the phase dependence. Thus with proper windowing, the variance of the frequency estimate is no more eclipsed by phase dependence, but it continues to decrease linearly with increasing SNR. The study covers both the cases of a single sinusoid and two sinusoids closely spaced in the frequency with the Pisarenko frequency estimator, MUSIC and principal component autoregressive frequency estimators. The trade-offs between the spectral broadening and the achieved minimum variance level due to the data window are analyzed in detail.Article Citation - WoS: 9Citation - Scopus: 15Subspace-Based Frequency Estimation of Sinusoidal Signals in Alpha-Stable Noise(Elsevier Ltd., 2002) Altınkaya, Mustafa Aziz; Deliç, Hakan; Sankur, Bülent; Anarım, EminIn the frequency estimation of sinusoidal signals observed in impulsive noise environments, techniques based on Gaussian noise assumption are unsuccessful. One possible way to find better estimates is to model the noise as an alpha-stable process and to use the fractional lower order statistics (FLOS) of the data to estimate the signal parameters. In this work, we propose a FLOS-based statistical average, the generalized covariation coefficient (GCC). The GCCs of multiple sinusoids for unity moment order in SαS noise attain the same form as the covariance expressions of multiple sinusoids in white Gaussian noise. The subspace-based frequency estimators FLOS-multiple signal classification (MUSIC) and FLOS-Bartlett are applied to the GCC matrix of the data. On the other hand, we show that the multiple sinusoids in SαS noise can also be modeled as a stable autoregressive moving average process approximated by a higher order stable autoregressive (AR) process. Using the GCCs of the data, we obtain FLOS versions of Tufts-Kumaresan (TK) and minimum norm (MN) estimators, which are based on the AR model. The simulation results show that techniques employing lower order statistics are superior to their second-order statistics (SOS)-based counterparts, especially when the noise exhibits a strong impulsive attitude. Among the estimators, FLOS-MUSIC shows a robust performance. It behaves comparably to MUSIC in non-impulsive noise environments, and both in impulsive and non-impulsive high-resolution scenarios. Furthermore, it offers a significant advantage at relatively high levels of impulsive noise contamination for distantly located sinusoidal frequencies.