Mehmet Belen

Dr. Mehmet Belen

orcid

Research Focus & Metrics

Academic Career Length

13Years (2014-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
15 September 2021 in IEEE Transactions on Antennas and Propagation
Verified
Journal Article
Closed access
Slawomir KozielSlawomir Koziel
Slawomir Koziel
NCNurullah Calik
Nurullah Calik
PMPeyman mahouti
Peyman mahouti
MBMehmet Belen
Mehmet Belen
The importance of surrogate modeling techniques has been gradually increasing in the design of antenna structures over the recent years. Perhaps the most important reason is a high cost of full-wave electromagnetic (EM) analysis of antenna systems. Although imperative in ensuring evaluation reliability, it entails considerable computational expenses. These are especially pronounced when carrying out EM-driven design tasks such as geometry parameter tuning or uncertainty quantification, both requiring repetitive simulations. Conducting some of the design procedures, e.g., global search or yield optimization, directly at the level of simulation models may be prohibitive. The use of fast replacement models (or surrogates) may alleviate these difficulties; yet, accurate modeling of antenna structures faces its own challenges. The two major obstacles are the curse of dimensionality, manifesting itself in a rapid growth of the number of training data samples necessary to render a reliable model (as a function of the number of antenna parameters) and high nonlinearity of antenna characteristics. Recently, the concept of performance-driven modeling has been introduced, where the modeling process is focused on a small region of the parameters’ space, which contains high-quality designs with respect to the considered performance figures. The most advanced variation in this class of methods is nested kriging, where both the model domain and the surrogate itself are constructed through kriging interpolation. Domain confinement is realized using a set of preoptimized reference designs and allows for significant improvement of the model predictive power while using a limited number of training data samples. In this work, the constrained modeling concept is coupled with a novel pyramidal deep regression network (PDRN) surrogate, which offers improved handling of highly nonlinear antenna responses. Three examples of microstrip antennas are used to demonstrate the advantages of constrained PDRN metamodels over the nested kriging surrogates with the (average) accuracy improved by a factor of 2 without increasing the training dataset cardinality.

10 May 2021 in IEEE Access
Verified
Journal Article
Open access
Slawomir KozielSlawomir Koziel
Slawomir Koziel
PMPeyman mahouti
Peyman mahouti
NCNurullah Calik
Nurullah Calik
MBMehmet Belen
Mehmet Belen
SSStanislaw Szczepanski
Stanislaw Szczepanski
Fast replacement models (or surrogates) have been widely applied in the recent years to accelerate simulation-driven design procedures in microwave engineering. The fundamental reason is a considerable-and often prohibitive-CPU cost of massive full-wave electromagnetic (EM) analyses related to solving common tasks such as parametric optimization or uncertainty quantification. The most popular class of surrogates are data-driven models, which are fast to evaluate, versatile, and easy to handle. Notwithstanding, the curse of dimensionality as well as the utility demands (e.g., so that the model covers sufficiently broad ranges of the system operating conditions), limit the applicability of conventional methods. A performance-driven modeling paradigm allows for mitigating these issue by focusing the surrogate setup process in a constrained domain encapsulating designs being of high quality w.r.t. the assumed figures of interest. The nested kriging framework capitalizing on this idea, renders the constrained surrogate using kriging interpolation, and has been shown to surpass traditional approaches. In pursuit of further accuracy improvements, this work incorporates the performance-driven concept into the fully-connected regression model (FRCM). The latter has been recently introduced in the context of frequency selective surfaces, and combined deep neural networks with Bayesian optimization, the latter employed to determine the network architecture and hyper-parameters. Using two examples of miniaturized microstrip couplers, our methodology is demonstrated to outperform both conventional modeling techniques and nested kriging, with reliable models constructed over multi-dimensional parameters spaces using just a few hundreds of samples.

8 March 2023 in IEEE Access
Verified
Journal Article
Open access
Slawomir KozielSlawomir Koziel
Slawomir Koziel
MBMehmet Belen
Mehmet Belen
AÇAlper Çalışkan
Alper Çalışkan
PMPeyman mahouti
Peyman mahouti
Reflectarrays (RAs) exhibit important advantages over conventional antenna arrays, especially in terms of realizing pencil-beam patterns without the employment of the feeding networks. Unfortunately, microstrip RA implementations feature narrow bandwidths, and are severely affected by losses. A considerably improved performance can be achieved for RAs involving grounded dielectric layers, which are also easy to manufacture using 3D printing technology. Regardless of the implementation details, a practical bottleneck of RA design is the necessity of independent adjustment of a large number of unit cells, which has to be carried out using full-wave electromagnetic (EM) simulation models to ensure reliability. The associated computational costs are extraordinary. A practical workaround is the incorporation of surrogate modeling methods; however, a construction of accurate metamodel requires a large number of training data samples. This letter introduces an alternative RA design approach, where the unit cells are adjusted using an inverse surrogate model established with a small number of anchor points, pre-optimized for the reference reflection phases. To ensure solution uniqueness, the anchor point optimization involves regularization, here, based on the minimum-volume condition for the unit cell. The presented approach reduces the computational cost of RA design to a few dozens of EM analyses of the cell. Several demonstration examples are provided, along with an experimental validation of the selected RA realization.

20 July 2022 in IEEE Transactions on Antennas and Propagation
Verified
Journal Article
Closed access
PMPeyman mahouti
Peyman mahouti
MBMehmet Belen
Mehmet Belen
NCNurullah Calik
Nurullah Calik
Slawomir KozielSlawomir Koziel
Slawomir Koziel
Reflectarrays (RAs) have been attracting considerable interest in the recent years due to their appealing features, in particular, a possibility of realizing pencil-beam radiation patterns, as in the phased arrays, but without the necessity of incorporating the feeding networks. These characteristics make them attractive solutions, among others, for satellite communications or mobile radar antennas. Notwithstanding, available microstrip implementations are inherently narrow-band, and heavily affected by conductor and surface wave losses. RAs based on grounded dielectric layers offer improved performance and flexibility in terms of shaping the phase reflection response. In either case, a large number of variables (induced by the need for independent adjustment of individual unit cell geometries), and the necessity of handling several requirements, make the design process of reflectarrays a challenging endeavor. In particular, RA optimization is extremely expensive when conducted at the level of EM simulation models, otherwise necessary to ensure reliability. A practical solution is surrogate-assisted design with the metamodels constructed for the RA unit elements. Unfortunately, conventional modeling methods require large numbers of training data samples to render accurate surrogates, which turns detrimental to the optimization process efficiency. This work proposes an alternative approach with the unit element representations constructed using deep learning with automated adjustment of the model architecture. As a result, design-ready surrogates can be constructed using only a few hundred samples, and the total RA optimization cost is reduced to only a handful of equivalent EM analyses of the entire array. Our approach is validated using an RA incorporating 3D pyramidal-shaped elements and favorably compared to benchmark techniques. Experimental verification of the obtained design is discussed as well.

26 October 2022 in IEEE Transactions on Antennas and Propagation
Verified
Journal Article
Closed access
Slawomir KozielSlawomir Koziel
Slawomir Koziel
NCNurullah Calik
Nurullah Calik
PMPeyman Mahouti
Peyman Mahouti
MBMehmet Belen
Mehmet Belen
The awareness and practical benefits of behavioral modeling methods have been steadily growing in the antenna engineering community over the last decade or so. Undoubtedly, the most important advantage thereof is a possibility of a dramatic reduction of computational expenses associated with computer-aided design procedures, especially those relying on full-wave electromagnetic (EM) simulations. In particular, the employment of fast replacement models (surrogates) allows for repetitive evaluations of the antenna structure at negligible cost, thereby accelerating processes such as parametric optimization, multi-criterial design, or uncertainty quantification. Notwithstanding, a construction of reliable data-driven surrogates is seriously hindered by the curse of dimensionality and the need for covering broad ranges of geometry/material parameters, which is imperative from the perspective of design utility. A recently proposed constrained modeling approach with knowledge-based stochastic determination of the model domain addresses this issue to a large extent and has been demonstrated to enable quasi-global modeling capability while maintaining a low setup cost. This work introduces a novel technique that capitalizes on the domain confinement paradigm and incorporates deep-learning-based regression modeling to facilitate handling of highly-nonlinear antenna characteristics. The presented framework is demonstrated using three microstrip antennas and favorably compared to several state-of-the-art techniques. The predictive power of our models reaches remarkable two percent of a relative RMS error (averaged over the considered antenna structures), which is a significant improvement over all benchmark methods.

14 January 2019 in International Journal of RF and Microwave Computer-Aided Engineering
Verified
Journal Article
Closed access
PDF file
HPHakan P. Partal
Hakan P. Partal
MBMehmet Belen
Mehmet Belen
SPSibel Zorlu Partal
Sibel Zorlu Partal