ABSTRACT. RF and Microwave active filters design by the use of grounded active inductor (AI) is here presented. The proposed single transistor AI emulates an inductor behaviour by implementing a passive variable phase and amplitude compensating network and amplifiers, forming a similar gyrator-C architecture allowing its use in LC filters architectures at high frequencies with high quality factor and reduced occupied chip area. The design method can be applied with success, in particular, for the design of bandpass filters with very high performances in terms integration and application from RF to Microwave frequency range achieving active filters with relatively high dynamic range, constant Q and also easy frequency tuning capability.

ABSTRACT. This paper explores the application of band pass filters with antiparallel configurations for obtaining dual band filters of compact sizes and various characteristics. The previously developed model for filters with antiparallel configuration is enhanced by replacing the ideal inductances with transmission lines. The characteristic impedance of the transmission lines is introduced as an additional independent parameter that affects the occurrences and the position of the parasitic passband at higher frequencies. This property is implemented within a program for electrical circuit simulation to obtain a model of dual band bandpass filter with tunable characteristics and simultaneous calculation of values for all components of an ideal circuit model. The obtained component values of the filter prototype can be scaled to an arbitrary microwave frequency.

ABSTRACT. The paper presents the comparison of full substrate integrated waveguide and rectangular waveguide surrogate model when calculating loss at Ka band. The design of the surrogate model itself, the waveguide transition to microstrip line, and the conversion of the surrogate model to the full model have been presented. The influence of the conductor and dielectric losses and the effect of surface roughness have been studied. It has been shown that the surrogate model reduces the numerical complexity of the full model as it requires more than eight times less unknowns which results in more then thirty five times faster simulations without any loss of the accuracy. For all of the cases considered the maximum difference between the loss values calculated using two models was 0.04 dB for a waveguide section of 3.5 wavelength at the middle-band frequency.

ABSTRACT. This paper presents new nonlinear crosstalk compensation technique in Multiple Input Multiple Output (MIMO) 4G Wireless Transmitters. It has been demonstrated that algorithm can suppress almost all distortion introduced by nonlinear crosstalk and Power Amplifier (PA) nonlinearity while significant complexity reduction compared with previously proposed Crossover Digital Predistortion (CO-DPD) technique is achieved. The technique is tested on 4x4 4G MIMO systems in the presence of large IQ imbalance.

ABSTRACT. This paper introduces an investigation of one type of position sensors. It is a resonant microstrip sensor with a resonator as an open stub controlled by a simple slot. It produces signal with frequency variation corresponding to covering of the slot in the ground plane. The sensing is performed by sliding the metal or dielectric plate over the slot. The structure is in planar technology, simple, low fabrication cost and compact. It is without multilayers, coupling, via holes, air bridges, discrete or active components.

ABSTRACT. The noise wave model provides relationships between the noise wave parameters and the noise parameters. However, since the noise wave model refers to the transistor intrinsic circuit, whose noise parameters are not directly measurable, the noise wave parameters are extracted based on the measured transistor noise parameters usually using time-consuming optimization procedures in circuit simulators. In this paper a new, faster and more efficient extraction approach based on artificial neural networks is proposed. The validation of the proposed approach is done in the case of the specific packaged HEMT based on AlGaAs/GaAs heterojunction by comparing the transistor noise parameters obtained using the extracted noise wave parameters with the measured noise parameters.

ABSTRACT. In this paper the design of the VHF Gysel 3 dB power divider/combiner is presented. The Gysel 3 dB power divider/combiner is designed as a balanced strip-line, for the frequency range from 150 MHz to 200 MHz, with central frequency at 175 MHz. The strip made of copper is placed inside the aluminum casing, filled with air. The Gysel 3 dB power divider/combiner has five ports, three of which are input/output ports, and the remaining two ports are connected with the external loads. Knowledge of the conditions, which must be satisfied by the Gysel, enables the acquisition of dimensions of the Gysel, through the process of computer simulation. In order to verify the project, values of the S-parameters of the simulated Gysel are compared to measured values of the S-parameters of the produced Gysel. Purpose of the Gysel 3 dB divider is to split an input signal into two equal outputs (equal by amplitude and by phase). In reverse direction, Gysel 3 dB divider works as a combiner, meaning it combines two in-phase signals into an output signal.

ABSTRACT. U ovom radu predstavljen je antenski niz sa cirkularnom polarizacijom namenjen radarskim sistemima koji rade u ISM opsegu 24-24.25 GHz, na centralnoj frekvenciji od 24.15 GHz. Predloženi antenski niz sastoji se 24 zasečene peč antene sa cirkularnom polarizacijom. Predloženi antenski niz projektovan je i testiran u WIPL-D softverskom okruženju. Rezultati su pokazali da predloženi niz ima dobre performanse, naime, elipticitet na željenom opsegu je manji od 2 dB, pojačanje je iznad 17 dB, dok je 3-dB širina glavnog snopa zračenja 5.2o.

ABSTRACT. U ovom radu izvešćemo približni analitički oblik za parametre rasejanja (transmisiju i refleksiju) za vod spregnut sa antisimetričnim rezonatorima -- slučaj u kome geometrija ne poseduje ravan simetrije, nego ostaje nepromenjena pri delovanju inverzije ili rotacije. Ovakav slučaj je zanimljiv zato što je moguće pobuditi dodatne rezonante modove, koji su nedostupni u simetričnim strukturama. Izvođenje se bazira na teoriji spregnutih modova, a dobijene formule biće primenjene na slučaj mikrostrip voda spregnutog sa split-ring rezonatorima, nepoznati parametri biće određeni fitovanjem i rezultati će biti upoređeni sa elektromagnetnim simulacijama.

ABSTRACT. Merenja elektromagnetske emisije ispitnog uređaja u frekvencijskom opsegu od 30 MHz do 200 MHz se u Tehničkom opitnom centru realizuje bikonusnom antenom. Pri tome, merni prostor, svojim refleksijama utiče na rezultate merenja. U cilju sagledavanja tih uticaja i unapređivanja merenja, modelovana je bikonusna antena. Do modela se došlo kroz faze, poređenjem rezultata simulacija i eksperimenata. Predmet ovog rada je prikaz tih simulacionih modela, kao i odgovarajućih rezultata simulacija i eksperimenata.

ABSTRACT. U ovom radu je prikazana štampana dipol antena zasnovana na modifikaciji Sierpinski trougaonog fraktala. Korišćenje fraktalnih geometrija za projektovanje antena omogućava projektovanje električno malih antena koje rade na više rezonantnih učestanosti. Korišćenje ove fraktalne geometrije ima za cilj podešavanje rezonantnih frekvencija fraktalne antene na visokim učestanostima promjenom toka struje po površini antene. Simulacijama su analizirani parametri rasijanja, raspodjela struje i dijagram zračenja. Eksperimentalnim mjerenjima parametara rasijanja potvrđeni su simulacioni rezultati. Pokazano je da se upotrebom fraktalnih geometrija mogu projektovati male antene koje zrače iste dijagrame zračenja na više različitih frekvencija uz očuvanje željenih karakteristika.

ABSTRACT. Генералисани модел повратног удара са путујућим струјним извором (ГТЦС модел) представља општи облик модела из којег могу да се изведу сви инжењерски модели повратног удара атмосферског пражњења, модели преносног вода као и модели путујућег струјног извора. У оквиру ГТЦС модела је уведена функција пражњења канала и израчуната за специјалне случајеве. У овом раду смо применили нову нумеричку методу за израчунавање функцијa пражњења. Предложени метод има велику тачност, врло је ефикасан и релативно је једноставан.

ABSTRACT. Kada se koristi kao deo postavke za merenje dijagrama zračenja antena, analizator mreža pogodno je kontrolisati spoljašnjim računarom. Savremenim analizatorima mreža spolja se može upravljati korišćenjem dva programska modela: COM (Component Object Model) i SCPI (Standard Commands for Programmable Instruments). COM koristi binarni protokol i efikasniji je od SCPI modela, koji koristi tekstualni protokol. U ovom radu prikazaćemo kako se koristi COM model i ilustrovaćemo rad sistema za merenje dijagrama zračenja.

ABSTRACT. U ovom radu data je analiza modifikovane peč antene sa bliskim višestrukim rezonansama koja se može upotrebiti kao osnovni element u nizovima sa proširenim frekvencijskim opsegom u odnosu na one sa klasičnim pečom. Modifikovani peč je znatno fleksibilniji za podešavanje u odnosu na standardni peč, naročito u pogledu prilagođenja.

ABSTRACT. Овај рад приказује дизајн холографске антене у опсегу од 57 - 64 GHz која користи унипланарни извор површинског таласа (лансер) који се простире у ТМ¬0 моду. Лансер се састоји од слот антене напајане копланарним таласоводом, директора и рефлектора у слот технологији. Димензије лансера су оптимизоване на учестаности од 60 GHz и он је коришћен као побуђивачки елеменат три холографке антене које се међусобно разликују по облику холограма. Употребљени су холограми са тракама једнаке дебљине (модел 1) са најдебљим тракама на крају холограма (модел 2) и са најдебљим тракама у средини холограма (модел 2). Најбоље укупне карактеристике показује модел 2, мада модел 3 има највеће појачање од око 17 dBi, али његов дијаграм зрачења има велике лобове. Променом фреквенције у опсегу од 57 - 64 GHz главни сноп антене се скенира за 160 у XY-равни, тј. у равни која која је у правцу простирања површинског таласа.

ABSTRACT. Nursing staff involved in interventional cardiology is considerably exposed to ionizing radiation given their operating positions during procedures and placement of structural shielding. Since placing and wearing eye lens dosimeter is encumbering for staff, Monte Carlo simulation provide us with easy way to get indication on what doses to the eye lens are. Eye lens doses were estimated for three X-ray tube projections (PA, LAO and RAO) and tube voltages ranging from 80 kV to 110 kV with different protective equipment setups. Simulations were carried out using MCNPX code

ABSTRACT. Proton irradiation of charge trapping flash memory cells was simulated using Monte Carlo method. Simulated memory cells included SONOS, SANOS, TANOS and three modifications of TANOS memory cell with different control gate material (TiN, WN and W). Proton energies were in ranges from 50 keV to 125 keV and 5 MeV to 50 MeV. Results are given in form of absorbed dose to memory cell and active layer, cross section for single event effects and threshold voltage shift.

ABSTRACT. A methodology for non-destructive radiological characterisation by in situ measurement was prepared to support the decommissioning planning effort for the RA shut down research reactor, and to manage existing radioactive waste. The geometry models of semiconductor HPGe and scintillation NaI detectors were developed for the MCNP-5 Monte Carlo code. Radiography was applied to ascertain the physical dimensions of these detectors. Scanning with a collimated source was utilised to ascertain the dead layer of HPGe detectors. At last, the MCNP-5 code was used to determine the total inactive dead layer of HPGe detectors, that are needed in order to obtain the minimum discrepancy between estimated and experimental efficiency. This paper describes all these steps, and summarises the strategy for space activity measurements. Example application was given for radiological characterisation of contaminated HEPA filters.

ABSTRACT. Measurement of tritium high-level activity in used heavy water is extremely important for further activities on decommissioning of the RA shut down research reactor. This paper describes a technique for measurements of tritium high-level activity, based on spectrometry of X-rays generated by interactions of beta particles emitted from tritium with constituent elements of surrounding materials (Beta Particles Induced X-ray Spectrometry). X-rays penetrate, in general, much deeper into materials than beta particles, increasing significantly the detectability of tritium. This technique has few advantages compared to other methods: intensity of X-rays does not depend on the chemical forms of tritium, detector is not contaminated by tritium, etc. A detector with thin NaI crystal and beryllium window is used as an X-rays spectrometer. Proposed measurement technique differs from existing ones by using simpler equipment and materials for experimental setup, and also, by using extensive application of Monte Carlo simulations instead of earlier prepared tritium standards to quantify the amount of tritium in the sample from the measured X-ray intensity. Application of proposed technique was given for sample with 1 ml of heavy water used in the RA nuclear reactor as coolant and moderator.

ABSTRACT. Neutron irradiation of titanium oxide memristors is studied through Monte Carlo simulations of neutron transport. Separate software packages are utilized for registering primary knock-on atom distribution, resulting from incident neutrons, and subsequent energy deposition of these displaced ions. Energy deposition mechanisms for secondary ions are compared at various incident neutron energies. Ranges of neutron energy for which radiation-induced changes in the oxide layer are expected to cause greatest alterations of memristor functionality are identified and linked to energy deposition pathways.

ABSTRACT. Model based on a track structure Monte Carlo simulation of low-energy electron transport in thin film materials is presented. It is a modification of previous models of its kind, and it takes into account all relevant interactions — elastic scattering, inelastic scattering and bremsstrahlung. Results obtained with the new model are compared to those from MCNP5 simulations, as well as to experimental data. It is demonstrated that the proposed model can be utilized for practical calculations. The model shows better agreement with experimental values of the backscatter coefficient than MCNP5, with no significant increase in computation time.

ABSTRACT. In this paper we select most probable reactions of alkali metal ions (Li+, Na+, K+) with dimethoxy ethane (DXE) molecule and its fragment ions are selected in order to obtain appropriate gas phase enthalpies of formation for the products. Calculated cross sections were compared with existing experimental results obtained by guided ion beam tandem mass spectrometry. Three body association reaction of ions with DXE for three different pressures is studied and compared to experimental results. The scattering cross sections set as a function of kinetic energy and transport parameters for Na+ in DXE gas as a function of E/N (E -electric field; N-gas density) were obtained by using the Monte Carlo technique.

ABSTRACT. In this article, the specific electrical resistance () and PTCR effect Er/Yb doped BaTiO3 ceramics were investigated. The content of additive Er2O3 and Yb2O3 in doped samples were ranged from 0.01 to 1.0 at% Er/Yb. The samples were prepared by a conventional solid state sintering procedure and sintered at 1320 and 1350C for 4 hours. For low dopants concentration (0.01 at% Er/Yb), SEM analysis shows abnormal grain growth with the average size range between 20-40 μm for samples doped with Er2O3 and from 30-50 μm for samples doped with Yb2O3. The increase of dopants concentration in samples causes decrease of average grain size, and for samples doped with 1.0 at% Er/Yb, grain size range between 3-20m for samples doped with Er2O3 and between 1-10 m for samples doped with Yb2O3. The specific electrical resistance were measured in temperature range from 25°C to 170°C at different frequencies, from 100Hz to 1MHz. The value of specific electrical resistance at room temperature range from 2.18∙104 Ωcm to 4.24∙104 Ωcm for samples doped with Er sintered at 1320°C and from 1.81∙104 Ωcm to 1.82∙104 Ωcm for samples sintered at 1350°C. For samples doped with 0.01 at% Yb sintered at 1320°C electrical resistivity was from 1.79∙104 Ωcm to 3.28∙104 Ωcm and from 1.81∙104 Ωcm to 1.82∙104 Ωcm for samples sintered at 1350°C. To a temperature of 120°C, resistance has a slight increase with increasing of temperature, but above this temperature the resistance rapidly increasing. The value of the specific electrical resistance decreases with increasing concentration of Er/Yb, to a concentration of 0.5 at% Er/Yb, and then resistance increases. Also, with increasing frequency, resistivity is lower for an order of magnitude.

ABSTRACT. The contemporary trends of ceramics applications, especially within advanced and electronics ceramics materials has a variety of important applications. It is commonly accepted that perovskite and similar ceramics have fractal morphology and can be the best expressed in the language of fractals. The source of fractality is threefold. First, ceramic grains have fractal shape of their surface. Second, there are so called “negative space” made of pores and intergranular space, also have fractal structure. Being extremely complex, the pore space plays an important role in microelectronic, PTC, piezoelectric and other phenomena. Third, there is process of Brownian--fractal motions inside the material during sintering in the form of flowing micro-particles –ions, atoms, electrons. These triple factors, in combination, make the microelectronic environment of very peculiar electro-static and dynamics microelectronic environment. The main issue, which we want to underline here is the basic property of fractal analytic technic. It is scale invariant property. It allows studying intergranular micro-capacity and micro-impedances as well as fractal components affecting on overall impedances distribution. Constructive theory of fractal sets explains possible approach in recognizing microcapacitors with fractal electrodes, seen as iterative process of interpolation which is compatible with the model of grains itself. Since intergranular permeability is taken as a function of working temperature, the correction factor also involves fractals. All of these new fractal frontiers opens a new era in future microelectronic processing and miniaturizations.

ABSTRACT. Solarne ćelije sa fotoosetljivim pigmentom su najbliže sistemu fotosinteze koji je do sada čovečanstvo postiglo. Njene karakteristike zavise od vrste pigmenta koji se nalazi u solarnoj ćeliji. U radu je obrađen kurkumin kao potencijalni novi pigment u solarnoj ćeliji i upoređen sa najkorišćenijim cijanidinom. Rezultati su pokazali da solarna ćelija sa kurkumom (kurkumin) daje veću efikasnost od solarne ćelije sa malinom (cijanidin), kao i da molekul kurkumina ima veću apsorpciju fotona u vidljivom delu spektra u odnosu na cijanidin. Kada su deprotonovani kurkumin i cijanidin adsorbovani za klaster titanijum dioksida, dobijeni raspored energetskih nivoa ukazuje na veću verovatnoću prelaska elektrona iz molekula u titanijum dioksid u slučaju kurkumina nego u slučaju cijanidina. Na osnovu dobijenih rezultata, zaključili smo da kurkumin ima bolje osobine kao fotoosetljivi pigment u solarnoj ćeliji.

ABSTRACT. У раду су приказана испитивања магнето-импедансног (МИ) ефекта траке метастабилне легуре Fe₇₂Cu₁V₄Si₁₅B₈ након одгревања на температурама од 300 ^ОC, 450 ^ОC, 500 ^ОC и 700 ^ОC. XRD анализа је показала да полазна легура садржи α-Fe(Si) кристалну фазу са величином кристалита од око 40 nm и метастабилну Fe₂₃B₆ фазу. ДТА анализом је уочено да на температури од око 470 ^ОC почињу структурне трансформације, а на 500 ^ОC се јавља егзотермни пик који указује на први ступањ кристализације. Узорак одгреван на 500 ^ОC показује повећање МИ односа и осетљивости у односу на неодгревану легуру. Узорак одгреван на температури од 700 ^ОC након завршеног другог ступња кристализације карактерише знатно смањење МИ односа.

ABSTRACT. Silicijum dioksid (SiO2) ima veliki broj uloga u procesu proizvodnje integrisanih kola, kao i uticaj na krakteristike komponenata. Deblina SiO2 slojeva kreće se od 1-2nm do 1-2μm. Proces oksidacije odvija se na visokim temperaturama (700-1300)°C. Silicijumska pločica se izlaže protoku kiseonika (suva oksidacija) ili vodenoj pari (vlažna oksidacija). Proces se može tačno modelirati linearno paraboličnim modelom (Deal – Grove model). Cilj ovog rada je da se korišćenjem Athena, procesnog simulatora, pokaže koncept procesa oksidacije i ispita uticaj promene različitih faktora na debljinu SiO2 sloja. Dobijeno je dobro slaganje rezultata dobijenih simulacijom i analitičkim putem.

ABSTRACT. In this paper, we address the Age of Artificial Intelligence, which we define as a distant future in which humans will reach the HLAI (Human Level of Artificial Intelligence) point. How humans will relate to fundamental concepts, such as freedom and truth. What will happen with today's basic human drivers, such as the death drive, anxiety and enthusiasm. Will they remain as basic ways of our entanglements with Being? We are convinced that the answering these questions today have full meaning not only for understanding the future, but primarily because of our understanding of modern human society and our openness toward new possibilities.

ABSTRACT. In this paper, we propose a new method for iris recognition based on methods of AI deep learning. With comparative analysis, we will show a parameter of the system (ROC) for authentication, obtained from the image of the iris (the result after the segmentation and normalization phase) and for traditional biometric templates (the result of the coding phase with the Gabor wavelet filter). In this way, we believe that it is possible to achieve a more effective system with better security performance in the systems based on iris authentication.

ABSTRACT. Teaching cryptology represents one of the most challenging tasks in contemporary engineering education. Apart from setting high requirement for students (i.e. required mathematical knowledge and skills), teachers are also required to invest a lot in each individual student. In the previous paper on this subject matter we presented a CrypTool extension that enables the interconnection of different simulation environments. In this paper, we present an intelligent, problem based learning based tutoring system that is built upon that extension. The system is capable of generating different (level) problems from cryptology domain, developing adequate solutions, evaluating students' solutions and reffering them to appropriate studying materials.

ABSTRACT. U radu su opisani koncepti i arhitektura konvolucijskih neuronskih mreža, čija primena je široko rasprostranjena. Značaj primene ove vrste neuronskih mreža je veliki, te su i istraživanja u ovoj oblasti od posebnog interesa. Postignuti su već značajni rezultati, te su napori oko sistematizacije u oblasti opravdani. U radu su predstavljeni važni rezultati istraživanja u oblasti, predložena je konzistentna terminologija i ukazano na izazove daljih istraživanja.