The paper describes a problem of computer simulation of critical phenomena in complex social systems on a petascale computing systems in frames of complex networks approach. The three-layer system of nested models of complex networks is proposed including aggregated analytical model to identify critical phenomena, detailed model of individualized network dynamics and model to adjust a topological structure of a complex network. The scalable parallel algorithm covering all layers of complex networks simulation is proposed. Performance of the algorithm is studied on different supercomputing systems. The issues of software and information infrastructure of complex networks simulation are discussed including organization of distributed calculations, crawling the data in social networks and results visualization. The applications of developed methods and technologies are considered including simulation of criminal networks disruption, fast rumors spreading in social networks, evolution of financial networks and epidemics spreading.

We study theoretically the  possibility of few-cycle short bipolar optical pulse compression and their transformation to unipolar pulses due to coherent interaction with resonance absorbing medium. It is shown that single-cycle pulse compression occurs when each half-wave starts to behave as an independent unipolar soliton. These solitons are attracted to each other under certain conditions, that leads to the emergence of single-cycle pulse of shorter duration. Numerical simulations revealed  3-5 times reduction of the pulse duration. The substantial absence of light loss in this scheme gives the possibility to create a multistage passive system of three resonance absorbers and results in a 125-time reduction of the pulse duration. Generation of unipolar pulses occurs when two powerful extremely short bipolar pulses propagate and collide in a dense resonant medium. In this case, as shown by numerical calculations, the mutual influence of oncoming solitons leads to the fact that some part of them is destroyed and another part is not. A high power unipolar soliton and low intensity bipolar optical ringing are observed in the medium output.

Subject of Research. The results of lidar combined scanning unit development for locating leaks of hydrocarbons are presented The unit enables to perform high-speed scanning of the investigated space in wide and narrow angle fields. Method. Scanning in a wide angular field is produced by one-line scanning path by means of the movable aluminum mirror with a frequency of 20Hz and amplitude of 20 degrees of swing. Narrowband scanning is performed along a spiral path by the deflector. The deflection of the beam is done by rotation of the optical wedges forming part of the deflector at an angle of ±50. The control function of the scanning node is performed by a specialized software product written in C# programming language. Main Results. This scanning unit allows scanning the investigated area at a distance of 50-100 m with spatial resolution at the level of 3 cm. The positioning accuracy of the laser beam in space is 15'. The developed scanning unit gives the possibility to browse the entire investigated area for the time not more than 1 ms at a rotation frequency of each wedge from 50 to 200 Hz. The problem of unambiguous definition of the beam geographical coordinates in space is solved at the software level according to the rotation angles of the mirrors and optical wedges. Lidar system coordinates are determined by means of GPS. Practical Relevance. Development results open the possibility for increasing the spatial resolution of scanning systems of a wide range of lidars and can provide high positioning accuracy of the laser beam in space.

Subject of Research. The paper deals with description of novel inkjet printing method for optical elements formed by structures based on TiO2 sol. The proposed approach presents effective way to obtain such optical nano objects as diffraction structures and transparent interference patterns. Methods. TiO2 nanoparticles were synthesized by hydrolysis of titanium isopropoxide (0.36 mol)  in deionized water (33.3 mol) under vigorous stirring and using nitric acid (0.04 mol) as a protonating agent. Viscosity was determined by Brookfield HA/HB viscometer, and surface tension - by Kyowa DY-700 tensiometer. Titanium dioxide inks were deposited on commercially available microembossed PET film with a thickness of 20 µm. To print titania ink Canon Pixma IP 2840 desktop office printer was used with a drop volume of 2 pL. The thickness of an inkjet TiO2 layer after drying in the air and removal of the solvents did not exceed 500 nm with a refractive index not less than 2.08 in the entire visible range.Main Results. The synthesis of aqueous TiO2 sol was used to obtain the ink with desirable rheological characteristics: viscosity and surface tension. The required rheology was regulated by controlling parameters of sol-gel transition in the system of aqueous titanium dioxide sol and by adding ethanol that affects the charge of double electrical layer of disperse phase. The reviled ratio of titanium dioxide sol and ethanol in the system gives such values of viscosity and surface tension that make this material convenient for inkjet printing. The coatings created by sol have a high refractive index in the entire visible range (not less than 2.08). We have shown that the deposition of optical transparent microstructures with diffraction effect has an ability to be applied on the transparent surfaces. The morphology of particles and the topology of printed structures were analyzed by optic and atomic-force microscopes.Practical Relevance. We have proposed the approach to obtain colorful interference patterns using one type of high refractive inks. The method opens new opportunities for “roll-to-roll” production technology of protected optical structures on the flexible polymer substrates.

Subject of Research.We present comparative results of the chirped Bragg gratings inscription efficiency in optical fiber of domestic production with and without low-temperature hydrogen loading. Method. Chirped fiber Bragg gratings inscription was made by the Talbot interferometer with chirped phase mask having a chirp rate of 2.3 nm/cm used for the laser beam amplitude separation. The excimer laser system Coherent COMPexPro 150T, working with the gas mixture KrF (248 nm), was used as the radiation source. In order to increase the UV photosensitivity, the optical fiber was placed in a chamber with hydrogen under a pressure of 10 MPa and kept there for 14 days at 40 °C. Main Results. The usage of the chirped phase mask in a Talbot interferometer scheme has made it possible to get a full width at half-maximum of the fiber Bragg grating reflection spectrum of 3.5 nm with induced diffraction structure length of 5 mm. By preliminary hydrogen loading of optical fiber the broad reflection spectrum fiber Bragg gratings with a reflectivity close to 100% has been inscribed. Practical Relevance. The resulting chirped fiber Bragg gratings can be used as dispersion compensators in optical fiber communications, as well as the reflective elements of distributed fiber-optic phase interferometric sensors.

Subject of Research. The problem of multiharmonic disturbance compensation for the class of linear time-invariant plants with known parameters and delay is considered.  Method. The disturbance is presented as unmeasurable output of linear autonomous model (exosystem) with known order and unknown parameters. The problem is resolved with the use of parametrized representation of disturbance designed by means of exosystem state observer and predictor of this state that finally enables applying certainty equivalence principle. In order to remove undesirable influence of delay a modified adaptation algorithm is created. The algorithm is based on augmentation of the plant state vector and generates advanced adjustable parameters for control. As distinct from widespread approaches, the proposed algorithm does not require identification of disturbance parameters and gives the possibility to remove such restrictions as adaptation gain margin and time delay margin. Main Results. Simulation results obtained in MATLAB/Simulink environment are presented to demonstrate the performance of proposed approach. Results illustrate the boundness of all signals in the closed-loop system and complete compensation of harmonic signal. It is shown that the proposed idea makes it possible to increase the adaptation gain for different delays without system stability loss. Practical Relevance. The algorithm of adaptive compensation  is recommended for the use in such problems as: the problem of control for active vibration protection devices wherein several dominating harmonics can be taken from the spectrum of vibration signal; the problems of control of robotics systems with periodical behavior; the problems of ship roll compensation; the problems of space plants control in the presence of uncontrollable rotation.

Subject of Study.The paper deals with novel research of ion exchange duration influence on spectral-luminescent properties of silver clusters formed in photo-thermo-refractive glass. Method. Photo-thermo-refractive matrix glass based on Na2O–Al2O3–ZnO–SiO2–F (% mol.) system doped with 0,002% mol. of Sb2O3 was synthesized for further research. Silver ions were introduced with low temperature ion exchange method. The glass samples were immersed in the mixture of sodium and silver nitrates 5AgNO3/95NaNO3 (% mol.) at the temperature of 320 °C. Ion exchange duration varied from 5 minutes to 21 hours. Luminescent silver clusters were formed in surface layers of photo-thermo-refractive glass by subsequent heat treatment at the temperature of 450 °C. Main Results. Embedding of silver ions in photo-thermo-refractive glass with ion exchange method led to long-wavelength shift of the UV edge of strong absorption. Location of the UV edge of strong absorption and emission peak of silver clusters depends on ion exchange duration and shifts to the greater wavelengthswith increasing the ion exchange process time. Quantum yield of luminescence decreases significantly according to Stern-Volmer equation with the rising of ion exchange duration. Practical Relevance. Research results can be used for developing white LEDs and down-convertors of solar radiation.  

By the use of high-resolution X-ray diffractometry and synchrotron radiation topography, the distribution of dislocations in AlN crystals grown on evaporating SiC substrates is studied. The growth of AlN layer in one process with the evaporation of SiC substrate gave the possibility to prevent the relaxation of thermal stresses through cracking of AlN during cooling the structure. The continuous 0.2-1.5 mm thick plates were used as the model objects for the study of dislocation structure near the AlN/SiC interface. Analysis of the broadening of the Bragg reflection peaks, the shape of scattering maps in reciprocal space and topographs showed that dislocations formed the mosaic structure, different from the distribution of threading dislocations in GaN epilayers. A theoretical model for misfit strain relaxation in the growing AlN layer is suggested. The results enable to clarify the dislocation processes during sublimation growth of industrial quality AlN crystals.  

Subject of Research.The paper deals with processes of formation and transformation of defects during multi-stage growth of thick gallium nitride layers with hydride vapor phase epitaxy on GaN/Al2O3 substrates with buried column pattern formed with the use of metal-organic vapor phase epitaxy. Methods. The growth of initial GaN layers was performed with the use of metal-organic vapor phase epitaxy. On the surface of the initial layers columns with the height of 800 nm were generated by means of ion etching. These columns were overgrown with 3-4 µm-thick GaN layers. On thus formed substrate multi-stage growth of GaN layers was performed with the use of hydride vapor-phase epitaxy. The total thickness of GaN layers was 100-1500 µm. The grown layers were studied by optical and electron microscopy and Raman spectroscopy. Main Results. Density of threading dislocations in the layers grown by hydride vapor-phase epitaxy was (3-6)·107 cm-2, that was one order of magnitude lower than in the used substrate, and two to three orders lower than dislocation density in typical GaN layers grown on commercial sapphire substrates. Raman spectroscopy data were indicative of low level of mechanical stress in the layers and their high structural uniformity. It was established that under multi-stage growth conditions, non-catastrophic cracks (those that do not cause sample destruction) are able to transform into macropores and appear to be an important st ructural element, serving to stress relaxation in the bulk of thick gallium nitride layers grown on foreign substrates. Practical Relevance. The results of the study can be used in the development of III-nitride heterostructures for optoelectronics and high-power and high-frequency microelectronics.

Subject of Research.The paper presents results of experimental study of measurement features in liquids by atomic force microscope to identify the best modes and buffered media as well as to find possible image artifacts and ways of their elimination. Method. The atomic force microscope Ntegra Aura (NT-MDT, Russia) with standard prism probe holder and liquid cell was used to carry out measurements in liquids. The calibration lattice TGQ1 (NT-MDT, Russia) was chosen as investigated structure with a fixed shape and height. Main Results. The research of probe functioning in specific pH liquids (distilled water, PBS - sodium phosphate buffer, Na2HPO4 - borate buffer, NaOH 0.1 M, NaOH 0.5 M) was carried out in contact and semi-contact modes. The optimal operating conditions and the best media for the liquid measurements were found. Comparison of atomic force microscopy data with the results of lattice study by scanning electron microscopy was performed. The features of the feedback system response in the «probe-surface» interaction were considered by the approach/retraction curves in the different environments. An artifact of image inversion was analyzed and recommendation for its elimination was provided. Practical Relevance. These studies reveal the possibility of fine alignment of research method for objects of organic and inorganic nature by atomic force microscopy in liquid media.

Subject of Research. We study speaker adaptation of deep neural network (DNN) acoustic models in automatic speech recognition systems. The aim of speaker adaptation techniques is to improve the accuracy of the speech recognition system for a particular speaker. Method. A novel method for training and adaptation of deep neural network acoustic models has been developed. It is based on using an auxiliary GMM (Gaussian Mixture Models) model and GMMD (GMM-derived) features. The principle advantage of the proposed GMMD features is the possibility of performing the adaptation of a DNN through the adaptation of the auxiliary GMM. In the proposed approach any methods for the adaptation of the auxiliary GMM can be used, hence, it provides a universal method for transferring adaptation algorithms developed for GMMs to DNN adaptation.Main Results. The effectiveness of the proposed approach was shown by means of one of the most common adaptation algorithms for GMM models – MAP (Maximum A Posteriori) adaptation. Different ways of integration of the proposed approach into state-of-the-art DNN architecture have been proposed and explored. Analysis of choosing the type of the auxiliary GMM model is given. Experimental results on the TED-LIUM corpus demonstrate that, in an unsupervised adaptation mode, the proposed adaptation technique can provide, approximately, a 11–18% relative word error reduction (WER) on different adaptation sets, compared to the speaker-independent DNN system built on conventional features, and a 3–6% relative WER reduction compared to the SAT-DNN trained on fMLLR adapted features.

Problem Statement. This paper describes the modeling of sonification system with possible types of wireless sensor network data. Fuzzy logic is used for the data-to-sound mapping. Methods. Devised sonification system includes input data model and sound synthesis core. It was created in Pure Data. For fuzzy output of mapped data the Fuzzy Logic Toolboxof MATLABwas used. Moreover, the system model has an ability to send data to the side application via UDP protocol. Results. We offer the method of timbre space organization for sonification system output and the following output of control sound characteristics depending on the type of input data. Practical Relevance. The offered approach of using fuzzy logic in sonification systems can be applied in development of new applications when the formalization of data-to-sound mapping is difficult and also complicated timbal space organization is required.

The paper deals with creation of integrated information system architecture capable of supporting management decisions using behavioral features. The paper considers the architecture of information decision support system for production system management. The behavioral feature is given to an information system, and it ensures extraction, processing of information, management decision-making with both automated and automatic modes of decision-making subsystem being permitted. Practical implementation of information system with behavior is based on service-oriented architecture: there is a set of independent services in the information system that provides data of its subsystems or data processing by separate application under the chosen variant of the problematic situation settlement. For creation of integrated information system with behavior we propose architecture including the following subsystems: data bus, subsystem for interaction with the integrated applications based on metadata, business process management subsystem, subsystem for the current state analysis of the enterprise and management decision-making, behavior training subsystem. For each problematic situation a separate logical layer service is created in Unified Service Bus handling problematic situations. This architecture reduces system information complexity due to the fact that with a constant amount of system elements the number of links decreases, since each layer provides communication center of responsibility for the resource with the services of corresponding applications. If a similar problematic situation occurs, its resolution is automatically removed from problem situation metamodel repository and business process metamodel of its settlement. In the business process performance commands are generated to the corresponding centers of responsibility to settle a problematic situation. 

The usage of various packet scheduling disciplines in computer networking devices as a mechanism to ensure the quality of service is described. Stages for selection of necessary parameters values of packet scheduling during VoIP-traffic transmission in computer networks are defined. VoIP-traffic was set as a research object because there are strict requirements of VoIP-applications to the network transmission parameters. With the aid of training and experimental simulation system the numerous experiments for parameters selection of the most common packet scheduling disciplines were carried out (FIFO, WFQ, non-preemptive priority queueing). The example that illustrates the ability to adjust the weighting coefficients of WFQ packet scheduling discipline is presented. Approximate analytical dependences are obtained and they will significantly reduce system administrators’ efforts to assess and modify the parameters of packet scheduling in network devices. A method of automating selection of the optimal packet scheduling discipline is formulated.

The new paradigm in the field of access control systems with fuzzy authorization is proposed. Let there is a set of objects in a single data transmissionnetwork. The goal is to develop dynamic authorization protocol based on correctness of presentation of events (news) occurred earlier in the network. We propose mathematical method that keeps compactly the history of events, neglects more distant and least-significant events, composes and verifies authorization data. The history of events is represented as vectors of numbers. Each vector is multiplied by several stochastic vectors. The result is known that if vectors of events are sparse, then by solving the problem of -optimization they can be restored with high accuracy. Results of experiments for vectors restoring have shown that the greater the number of stochastic vectors is, the better accuracy of restored vectors is observed. It has been established that the largest absolute components are restored earlier. Access control system with the proposed dynamic authorization method enables to compute fuzzy confidence coefficients in networks with frequently changing set of participants, mesh-networks, multi-agent systems. 

Subject of Research. The paper considers the problem of automatic term extraction from natural language texts (text mining). One of the first-priority problems in this topic is creation of domain thesaurus. Some well approved methods of terms extraction exist for alphabetic languages, for instance, the latent semantic analysis. Applying of these methods for hieroglyphic texts is challenged because of missing blanks between words. The sentences segmentation task in hieroglyphic languages is usually solved by dictionaries or by statistical methods, particularly, by means of a mutual information approach. Methods of sentences segmentation, as methods of terms extraction, separately, do not reach 100 percent accuracy and fullness, and their consistent applying just increases a number of errors.  The aim of this work is improving the fullness and accuracy of domain terms extraction from hieroglyphic texts. Method.The proposed method lies in detection of repeating two, three or four symbol sequences in each sentence and correlation of occurrence frequencies for these sequences in domain and contrast documents collection. According to research carried out it was stated that a trivial ranging of all possible symbol sequences enables to extract satisfactory only frequently using terms. Filtering of symbol sequences by their ratio of frequencies in the domain and contrast collection gave the possibility to extract reliably frequently used terms and find satisfactory rare domain terms. Some results of terms extraction for the “Network technologies” domain from a Chinese text are presented in this paper. A set of articles from the newspaper “Rénmín Rìbào” was used as a contrast collection and some satisfactory results were obtained.

Subject of Research. The paper deals with the methods of efficiency improving for piezoelectric vibration sensors used in aircraft industry to control the level of vibration of gas turbine engines. The study looks into the matter of surface microgeometry effect of the vibro sensor part on its transverse sensitivity ratio. Measures are proposed to improve the sensor performance without cost supplement by optimization of the functional surface microgeometry. Method. A method for determination of the best possible surface microgeometry within the specific production conditions is shown. Also, a method for microgeometry estimation of the functional surfaces using graphical criteria is used. Taguchi method is used for design of experiment for functional surfaces machining. The use of this method reduces significantly the number of experiments without validity loss. Main Results. The relationship between technological factors of manufacturing the vibration sensor parts and its sensitivity has been found out. The optimal surface machining methods and process conditions for parts ensuring the best possible sensitivity have been determined. Practical Relevance. Research results can be used by instrument-making companies to improve the process of piezoelectric vibration sensor design and manufacturing.

The paper deals with mathematical model presentation of cash flows in project funding. We describe different types of expenses linked to university project activities. Problems of project budgeting that contribute most uncertainty have been revealed. As an example of the model implementation we consider calculation of vacation allowance expenses for project participants. We define problems of forecast for funds reservation: calculation based on methodology established by the Ministry of Education and Science calculation according to the vacation schedule and prediction of the most probable amount. A stochastic model for vacation allowance expenses has been developed. We have proposed methods and solution of the problems that increase the accuracy of forecasting for funds reservation based on 2015 data.

Subject of Research. We proposea computational tool to evaluate the probability-time characteristics of info-communication network as the quality indicators in the selection of effective design solutions. For the assessment of probability-time characteristics of info-communications networks we use simulation that makes provision for the dynamic and stochastic aspects of their functioning.We recommend an approach to reduce simulation model dimensionality based on the decomposition of info-communication network into sets of virtualchannels. Methods. Topologically complete info-communication network is a virtual channel model that connects the source and destination, and the probability equivalent of not considered network part is represented as a generator of external flows for taking into account the interaction peculiarities of considered virtual channel flow with flows of other routes. One and the same virtual channel model for representation of different segments of info-communication network differs only in parameters. Main Results. It is shown that the object approach application gives the possibility to build structures and variations of the virtual channel parameters corresponding to real segments of info-communication network. Convergence of the analytical modeling results and simulation is demonstrated. Thus, the use of the virtual channel simulation model as a computational tool is possible to evaluate the probability-time characteristics of the info-communications network. Practical Relevance. Virtual channel model can be useful as a tool for planning and forecasting of computing resources indicators of the info-communication network quality.

The paper considers the challenges of providing mechanical systems quality at the design stages. The engineering solution choice in the design of mechanical systems has a significant impact on the product quality. It is noted that in practice a developer often resorts to an expert estimation when solving choice problems. Known methods of quality forecasting are based mainly on statistical data. This leads to limited application possibilities for these methods in practice for quality forecasting in the development process of original single-piece mechanical products with the new engineering solutions built in them. Proposed method is based on the relative qualitative and quantitative evaluation of different design options for developed products aimed at reasonable choosing of optimal engineering solution from a number of alternative ones. The final scope of proposed and considered method is quality improvement of the mechanical products being developed. A distinguishing feature of the method is determination of the main quality indicator, the basic quality indicators, and the usage of reference basic technical solution to estimate and compare development solutions. The base for qualitative assessment of options is modified expert evaluation method. Quantitative assessment of options is based on the analysis of constructive and technological parameters that influence the main quality indicator and basic quality indicators. The proposed method is examined on the example of operating mechanism gears, optical-mechanical systems and automatic gripper of a cluster fuel assembly. The proposed technique enables to increase the product development quality due to lowering the developer subjectiveness, increase of information analysis quality and the usage of staff engineering potential concerning the aspects of engineering solution choice from a number of alternative ones.  The algorithm for proposed methodology is described, its distinguishing features are highlighted, and the conclusions about advisability of this methodology application in practice are made.

The paper describes our results on antifungal effect of nanocomposite material. The nanocomposite is based on acrylates and functional nanoparticles composition. The study of material activity was carried out by the agar diffusion method. Candida albicans (C. albicans) strain of fungi was used as a test object. The antifungal effect of polymer matrix is shown. Zinc oxide nanoparticles increase the antifungal effect. 

Subject of Research. We study organization of machine-to-machine exchange over aggregated channels divided into groups. For requests on packages transfer a common queue can be arranged. As a result of multiple access for each request at the same time the resources of all channels in the group are provided. A package copy is transmitted through each channel. Methodology of Research. Study of machine-to-machine exchange effectiveness is based on analytical modeling. Due to the identical character of channels and the simultaneity of their provision to a packet transmission, it is assumed that all copies will be delivered to the recipient at the same time, although transmission errors or packet loss are possible. That proves the channel group representation as a one servicing device. Main Results. We have shown the efficiency of redundant transmission organization of copies of packets over aggregated channels divided into groups. We have defined the boundaries of appropriate application of redundant transmission of packets, time-critical to their unmistakable delivery over the aggregated channels divided into groups. Practical Relevance. The proposed models can be used in selection of decisions aimed at the organization of redundant transmission over the aggregated channels.

The effect of initial network topology on a temporal network dynamics is studied. An example of interbank exposures network is considered. It is modeled with a graph, where banks are represented by nodes and interbank lending is represented by edges. The dynamical processes in аtemporal network are defined by state changes of nodes and lie in edges and nodes addition and deletion in a graph, and modification of node states contribute to network evolution. The algorithm of network modification over the whole evolution period is fixed. We present parameters of random, scale free and small world generative models corresponding to different simulation results with fixed modification algorithms. The influence of initial graph topologies on temporal network dynamics is demonstrated. The results obtained give the possibility to assess time interval before the attainment of unstable topology state, and to estimate an optimal topology for the transition to a steady state under fixed modification algorithms.

The paper presents two methods of predicting the peak of influenza epidemics using population-based mathematical models: Baroyan-Rvachev and modified Kermack-McKendrick model, proposed by the authors. We perform the comparison of the prediction accuracy of time and the value of epidemics peaks on long-term data of ARI incidence in the city of St. Petersburg. The methodology of comparison is based on three criteria of accuracy conventionally named as "square", "vertical stripe" and "horizontal stripe", and two variants of the model parameters estimation. In the first variant we calibrate the model on the data of the first city impacted by the epidemic, and use these parameters in the future for the other cities, that allows taking into account the spatial characteristics of the epidemic in the country. In the second case, we only use historical data available at the time of the prediction for a given city. The advantage of this approach is the lack of need for additional, not always available, external data to predict the epidemic. The results of test calculations have demonstrated that the first method shows good results in the case of significant delays between the peaks of epidemics in different cities. If the outbreak in St. Petersburg started soon after the registration of the first outbreaks in the other cities of the Russian Federation, the second method shows comparable results to an accuracy of 90% to predict the peak of the epidemic. In most cases, it is sufficient for the use of the results of calculations for planning antiviral activities. The lead time of the peak prediction is still at a relatively low level, that seems to be associated with a variety of patterns of virus spread and permanent changes in transport communications within the country.