A study on transmit beamforming and noisy autoregressive modeling problems

Download

index.pdf

Date

2023-1

Author

Çayır, Ömer

Metadata

Show full item record

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Item Usage Stats

149
views

114
downloads

This thesis presents algorithms for two problems in statistical signal processing. The first problem is the transmit beamformer design under the peak-to-average power ratio (PAPR) constraint. With the aim of establishing a trade-off between the power efficiency (maximizing the average transmitted power in the main lobe) and other metrics, such as the fluctuation of power in the main lobe and the peak-sidelobe level, how the PAPR constraint affects the design problem is examined. In general, unimodular weights, which have a constant magnitude and therefore ensure the lowest possible PAPR value, are used to maximize the average transmitted power at the expense of other performance metrics. It is empirically shown that even a minor relaxation of the design problem from the lowest PAPR condition leads to a significant improvement in performance metrics at a negligible loss in power efficiency. A solution based on the alternating direction method of multipliers (ADMM) is provided to achieve the trade-off between the performance metrics. Moreover, a consensus ADMM-based solution is presented for the equivalent problem in consensus form. The proposed solutions can be used for both narrowband and wideband beamformers. The second problem is the maximum likelihood autoregressive (AR) model parameter estimation from the independent snapshots observed under additive white Gaussian noise. In addition to the AR model parameters, the measurement noise variance is included among the unknowns of the problem to develop a general solution covering several special cases, such as the case of known noise variance, noise-free snapshots, and the single snapshot operation. The presented solution is based on the expectation-maximization method, which is formulated by assigning the noise-free snapshots as the missing data. An approximate version of the suggested method, at a significantly reduced computational load with virtually no loss of performance, is also provided.

Subject Keywords

Transmit beamforming, Peak-to-average power ratio, PAPR, Alternating direction method of multipliers, ADMM, Autoregressive process, Autoregressive model parameter estimation, Multiple snapshots, Expectation-maximization

URI

https://hdl.handle.net/11511/101951

Collections

Graduate School of Natural and Applied Sciences, Thesis

Suggestions

OpenMETU
Core

Comparative study of attitude control methods based on Euler angles, quaternions, angle-axis pairs and orientation matrices Özgören, Mustafa Kemal (2019-03-01) This paper presents a comparative study about the attitude control methods based on four commonly used error indicators, namely the triad of 3-2-1 deviational Euler angles, the error quaternion, the deviational angle-axis pair and the orientation error matrix. These error indicators are used here with the same backstepping control law to have a common basis of comparison. This control law makes the controller track a restoring angular velocity generated here specifically for each error indicator. This compa...
On the reduction of Gaussian inverse Wishart mixtures Granström, Karl; Orguner, Umut (2012-09-12) This paper presents an algorithm for reduction of Gaussian inverse Wishart mixtures. Sums of an arbitrary number of mixture components are approximated with single components by analytically minimizing the Kullback-Leibler divergence. The Kullback-Leibler difference is used as a criterion for deciding whether or not two components should be merged, and a simple reduction algorithm is given. The reduction algorithm is tested in simulation examples in both one and two dimensions. The results presented in the ...
A linear approximation for training Recurrent Random Neural Networks Halıcı, Uğur (1998-01-01) In this paper, a linear approximation for Gelenbe's Learning Algorithm developed for training Recurrent Random Neural Networks (RRNN) is proposed. Gelenbe's learning algorithm uses gradient descent of a quadratic error function in which the main computational effort is for obtaining the inverse of an n-by-n matrix. In this paper, the inverse of this matrix is approximated with a linear term and the efficiency of the approximated algorithm is examined when RRNN is trained as autoassociative memory.
A modified algorithm for peer-to-peer security Akleylek, Sedat; Emmungil, Levent; NURİYEV, URFAT (2007-01-01) In this paper we present the steganographic approach to peer-to-peer systems with a modified algorithm. This gives the user a very high level of protection against being compelled to disclose its contents. Even the realization of the quantum computer cannot solve NP-hard problem in a polynomial time, a modified algorithm with steganographic use depending on Knapsack problem may make peer-to-peer systems secure.
Investigation of Stationarity for Graph Time Series Data Sets Güneyi, Eylem Tuğçe; Vural, Elif (2021-01-07) Graphs permit the analysis of the relationships in complex data sets effectively. Stationarity is a feature that facilitates the analysis and processing of random time signals. Since graphs have an irregular structure, the definition of classical stationarity does not apply to graphs. In this study, we study how stationarity is defined for graph random processes and examine the validity of the stationarity assumption with experiments on synthetic and real data sets.

Citation Formats

Ö. Çayır, “A study on transmit beamforming and noisy autoregressive modeling problems,” Ph.D. - Doctoral Program, Middle East Technical University, 2023.