Ahmet Kizilay

A Neural Network Solution for Identification and Classification of Cylindrical Targets above Perfectly Conducting Flat Surfaces

This paper evaluates the radar target identification and classification performance of neural networks. A set of features are derived from scattered fields calculated by using the image technique formulation and Moment Method (MoM). An RBF (Radial Basis Function) networkthat utilizes the feature set is proposed for target identification and classification. The database contains a finite number of samples of cylindrical targets at certain angles. A portion of the database is used to train the networkand the rest is used to test the performance of the neural networkfor target identification and classification. This workaims to find the heights measured from the surface and radiuses of the targets for identification of targets and determine the right target for classification of targets from the scattered field values.

ON THE TARGET CLASSIFICATION THROUGH WAVELET-COMPRESSED SCATTERED ULTRAWIDE-BAND ELECTRIC FIELD DATA AND ROC ANALYSIS

Abstract—This paper’s aim is to classify cylindrical targets from their ultrawide-band radar returns. To calculate the radar returns, image technique formulation is used to obtain the Electric Field Integral Equations (EFIEs). Then, the EFIEs are solved numerically by Method of Moment (MoM). Because of wide frequency range of the ultrawide-band radar signal, the database to be used for target classification becomes very large. To deal with this problem and to provide robustness, wavelet transform is utilized. Application of wavelet transform significantly reduces the size of the database. The coefficients obtained by wavelet transform are used as the inputs of the artificial neural networks (ANNs). Then, the actual performances of the networks are investigated by Receiver Operating Characteristic (ROC) analysis.

Computation of the scattered fields from an overfilled cavity embedded in a perfectly conducting ground plane

A decomposition method is applied to solve the electromagnetic scattering from an overfilled cavity embedded in a perfectly electric conducting (PEC) ground plane. The surface equivalence principle is used with the decomposition method to obtain a set of electric field integral equations (EFIEs) for the currents on the overfilled cavity surfaces and the perturbation current on the portion of PEC ground plane surface near the cavity. The Method of Moments (MoM) is used to solve the EFIEs in the frequency domain to obtain the scattered electric field. The method is presented with examples for TMz polarization and the results are compared with various other solution methods.

High resolution signal processing techniques for millimeter wave short range surveillance radar

In this paper, we investigated adaptive signal processing algorithms for high-resolution imaging of short range targets by 24 GHz millimeter wave (MMW) surveillance radar system. For test measurements, a step frequency continuous wave (SFCW) based Synthetic Aperture Radar (SAR) imaging system has been developed by using Anritsu MS4644A Vector Network Analyzer (VNA) and A-scan radar data were collected to demonstrate B-scan target images. We applied SAR processing and MUSIC (Multiple Signal Classification) algorithm for raw data to enhance the cross-range and range resolutions, respectively. The fusion of SAR and MUSIC algorithms is proposed to obtain improved image quality for better target identification and false alarm rate performance.

Front-end design for Ka band mm-Wave radar

In this paper front-end design for a Ka band milimeter wave (MMW) radar which consists of an antenna, a low noise amplifier (LNA) and a band pass filter is presented. The operation frequency of the designed system is between 24-25 GHz in Ka band. A high gain axially displaced elliptical (ADE) dual reflector antenna is employed on the antenna structure. ADE sub-reflector with 5 cm diameter is illuminated by a feeder horn and a main reflector 30 cm diameter parabola focuses incoming waves from the ADE sub-reflector. According to the simulation results narrow half power beam width (HPBW=30) and high gain (G=35 dBi) are obtained with good efficiency (%58). An HJ-FET that has low noise figure (NF<;1 dB) and high gain (>13 dB) is utilized to design a LNA. Double transistors are connected as cascaded to achieve higher transducer gain (Gt>19 dB). Matching circuits and feeder resonators are designed by microstrip lines to obtain low input and output VSWR (Vin<;2.1, Vout<;2.1). A microstrip band pass filter (BPF) is designed to receive required signals and to suppress other bands. The BPF is formed by combination of a radial stub low pass filter (LPF) and a short stub high pass filter. Low insertion loss (S21>-2.5dB) and low return loss (S11<;-15 dB) are aimed to take signal as lossless as at pass band. The simulated designs are manufactured and measured. It is seen that there are good agreements between measurement and simulation results.

Millimeter wave short range radar system design

In this paper, a 24 GHz millimeter wave (mm-wave) short range radar system is presented. In the designed radar system, there is one transmitter and one receiver channel. The transmitter channel has +9 dBm output power and 24 GHz output carrier signal is modulated with 20 ns square wave by SPDT switch. At the receiver channel, there are two LNAs, which have totally 40 dB gain, a microstrip band pass filter (BPF) and a double balanced IQ mixer. The designed radar system is measured using an oscilloscope at output of IF channel. According to the results, it is seemed that 200 mV and 50 mV reflection levels are received from metal and human target, respectively, at 2 meters without IF amplifier. ADE reflector antennas and horn antennas are utilized for transmitting and receiving.

Calculation of TM z scattering from a conductive object buried in a grounded lossy medium

The problem of electromagnetic scattering from a cylindrical object of arbitrary cross-section buried in a lossy-grounded dielectric medium is solved using a new numerical solution method. A set of electric field integral equations (EFIEs) for the currents on the object and the portion of the surfaces most strongly interacting with the object are obtained using the surface equivalence principle and a decomposition method. The Method of Moment (MoM) is used to solve the EFIEs in the frequency domain to obtain the scattered electric field.

A horn antenna array design to destroy cancerous tissues by non-destructive hyperthermia

In this paper, a novel approach to near field regional focusing for non-destructive hyperthermia is presented. It is aimed to use near-field patterns of antenna arrays for non-destructive treatment by heating of cancerous tissues of the body without damaging normal tissues. Initially, near-field radiation pattern of the horn antenna array placed on the body surface is obtained with a fast and reliable method of Analytical Regularization Method (ARM) in vacuum. At the same time, this pattern of the horn antenna array placed on the body surface is simulated in layered lossy environments (i.e., skin, fat, bone, tumor tissue) by Finite Integral Method (FIM). On the next studies, a hybrid method will be constructed between ARM and Method of Moments (MoM) and near-field patterns will be focused on the tumor using of optimization algorithm. In this manner, heat destruction of tumor tissue can be achieved without damaging the healthy parts of the organ. Finally, non-destructive hyperthermia performance of the designed dual-band horn antenna array system result of the theoretical calculations will be tested on artificial organ models.

24 GHz short range radar system measurements for synthetic aperture radar imaging

In this paper, 24 GHz short-range radar measurements and synthetic aperture radar (SAR) imaging results are presented. The designed radar system has one transmitter and one receiver channel and 24 GHz output carrier signal is modulated with 50 MHz square wave at the transmitter channel. In the measurement, horn antennas are used for synthetic radar imaging and A-scan data were obtained with 90 m range. We applied SAR imaging algorithm in order to improve the image resolution to raw B-scan data. Raw and processed data are demonstrated comparatively for target detection performance.

Scattering from an object behind multi-layer periodic gratings of cylindrical shapes

Abstract The problem of electromagnetic scattering from a perfect electric conductor (PEC) object of arbitrary cross-section behind multilayer of periodic gratings of circular cylinders is solved by a hybrid method combining the Method of Moments (MoM), decomposition, and the Physical Optics methods. The periodic grating’s periodic cell consists of multiple layers of lossy-dielectric scattering elements situated in free space. The PEC object under the periodic gratings assumed to be electrically large. A set of electric-field integral equations (EFIEs) for the perturbation currents both on the portion of the PEC object and the periodic gratings most strongly interacting with each other obtained using the surface equivalence principle and a decomposition method. Then, MoM is used to solve the EFIEs in the frequency domain to obtain the scattered electric field.