Theory and Engineering of Complex Systems and Dependability

The paper presents an attempt to anomaly detection in network traffic using statistical models with long memory. Tests with the GPH estimator were used to check if the analysed time series have the long-memory property. The tests were performed for three statistical models known as ARFIMA, FIGARCH and HAR-RV. Optimal selection of model parameters was based on a compromise between the model’s coherence and the size of the estimation error.

The paper concerns the risk assessment and management methodology in critical infrastructures. At the beginning a review is performed of the state of the art, regulations, best practices, EU projects, and other relevant documents. On this basis a set of the most preferable features of a CI risk management tool is identified. These features allow to specify basic requirements for the risk management tool. As the core of the solution is the bow-tie model. A risk register is proposed as an inventory of the hazardous events, along with other data structures for hazards/threats, vulnerabilities, consequences, and barriers. Risk factors and results measures, i.e. likelihood and consequences measures as well as a risk matrix are discussed. Next, a new concept is proposed how to integrate different bow-tie models through internal and external dependencies. These requirements can be implemented on the available software platform for further experiments and validation.

It is important to be able to predict if a module or a class or a method is faulty, or not. Such predictions can be used to target improvement efforts to those modules or classes that need it the most. We investigated the classification process (deciding if an element is faulty or not) in which the set of software metrics is used and examined several data mining algorithms. We conducted an experiment in which ten open source projects were evaluated by ten chosen metrics. The data concerning defects were extracted from the repository of the control version system. For each project two versions of code were used in the classification process. In this study the results of two algorithms i.e.

k

- NN and decision trees used in the classification process are presented.

The spatial interactions by definition represent a movement of people, freight or information between an origin and a destination. It is a relationship between transport demand and supply expressed over a geographical space. This method has a very particular importance in the transport geography, and relates how to estimate flows between locations, since these flows, known as spatial interactions, aim to evaluate the demand for transport services existing or potential. In this paper, the case study concerns the economic exchange in the Constantine city (which is a commune of north-eastern of Algeria) between the main towns (El-Khroub, Ain-Smara and Hamma-Bouziane) geographically belonging to the Constantine city. The gravity model is used, and it is the most common formulation of the spatial interaction method; it uses a similar formulation of Newton’s law of gravity.

Axiomatics for multivalued dependencies in table databases and axiomatics for functional and multivalued dependencies are reviewed; the completeness of these axiomatics is established in terms of coincidence of syntactic and semantic consequence relations; the completeness criteria for these axiomatic systems are formulated in terms of cardinalities (1) of the universal domain

D

, which is considering in interpretations, and (2) the scheme

R

, which is a parameter of all constructions, because only the tables which attributes belong to this scheme

R

are considering.

The results obtained in this paper and developed mathematical technique can be used for algorithmic support of normalization in table databases.

In this paper there are discussed the problems associated with yearly introduced by authors the class of dynamical systems, which have occurred from traffic. But now these dynamical systems become one of indicators of rational and irrational numbers connected with computer sciences basics.

Some results of computer experiments are presented. Also hypothesis and future works are provided.

A distributed information system for business processes of road maintenance is presented in the paper. The architecture of developed system is the type of client-server system with terminal devices such as smart phones with the Android and IOS Operating Systems. The system has been called “Server – Smartphone – Student – Receiver (SSSR) – Road”, and it gives possibility to fulfill the optimal scheduling, local routing for services movements and maintenance organization on the road traffic network by communications technology. Problems of dependability and system reliability are essentially developing on the accuracy of objects positioning. So methods of accuracy improving are provided in the paper.

The paper introduces a new approach to weather forecasting. Overall prediction process consisted of processing big data, turning processed data to visualization, and later this visualization has been used for enhancing forecasting methods using artificial neural networks. The following assumptions are proved: data visualization gives additional interpretation possibilities, it is possible to enhance weather forecasting by data visualization, neural networks can be used for visual weather data analysis, neural networks can be used for climate changes prediction.

As the functional complexity of the malicious software increases, their analyses faces new problems. The paper presents these aspects in the context of automatic analyses of Internet threats observed with the HoneyPot technology. The problems were identified based on the experience gained from the analyses of exploits and malware using the dedicated infrastructure deployed in the network of the Institute of Computer Science at Warsaw University of Technology. They are discussed on the background of the real-life case of a recent worm targeting Network Attached Storage (NAS) devices vulnerability. The paper describes the methodology and data analysis supporting systems as well as the concept of general and custom HoneyPots used in the research.

Web based information systems are exposed to various faults during their lifetime (originating in the hardware, in the software or stemming from security vulnerabilities). Service reconfiguration strategies are used to improve their resilience. A risk assessment approach is proposed to analyze the vulnerabilities of the system with reconfiguration. The proposed technique involves assessment of likelihood and consequences of occurrence of various fault combinations handled by the reconfiguration strategies.

The expected increase in transports of people and goods across Europe will aggravate the problems related to traffic congestion, accidents and pollution. As new road infrastructure alone would not solve such problems, Intelligent Transportation Systems (ITS) has been considered as new initiatives. Due to the complexity of behaviors, novel methods and tools for the requirements engineering, correct-by-construction design, dependability, product variability and lifecycle management become also necessary. This chapter presents an ontology-based approach to safety management in Cooperative ITS (C-ITS), primarily in an automotive context. This approach is supposed to lay the way for all aspects of ITS safety management, from simulation and design, over run-time risk assessment and diagnostics. It provides the support for ontology driven ITS development and its formal information model. Results of approach validation in

CarMaker

are also given in this Chapter. The approach is a result of research activities made in the framework of Swedish research initiative, referred to as SARMITS (Systematic Approach to Risk Management in ITS Context).

The aim of this paper was to propose an efficient algorithm for determining the best route for couriers’ in urban traffic. The best route is defined as the order in which the parcels should be delivered. The specification of couriers’ work may result in different criteria being considered optimal, including but not limited to: travelled distance, elapsed time, combination of preceding.

For this reason, the optimality criteria is defined and proposed solution is verified against those optimization targets.

The scope of this paper consists of researching modern methods of solving TSP. In addition to that, a simulation environment that allows to conduct experiments was developed. Secondly, a selected portion of methods described in literature was implemented to form a baseline for benchmarking proposed solution. Furthermore, authorial solutions are implemented and the results of running those are compared against baseline results. A best route is considered in terms of parcel delivery order and not finding the shortest path between subsequent delivery addresses. Consequently the routing problem (finding the shortest path) is beyond the scope of this paper.

In test-driven development, basic tests are prepared for a piece of program before its coding. Many short development cycles are repeated within the process, requiring a quick response of the prepared tests and the tested code extract. Mutation testing, used for evaluation and development of test cases, usually takes a considerable time to obtain a valuable test assessment. We discuss combination of these techniques in one development process. The presented ideas are implemented in VisualMutator – an extension of Visual Studio. The tool supports selected standard and object-oriented mutations of C# programs. Mutations are introduced at the level of the Common Intermediate Language of .NET. A program or its selected methods can be mutated in a batch process or during interactive sessions.

The paper includes a proposal of a new algorithm for Computer Aided Design (CAD) of complex system with higher degree of dependability. Optimal scheduling of processes and optimal resources partition are basic problems in this algorithm. The following criteria of optimality are considered: costs of system implementation, its operating speed and power consumption. Presented the CAD algorithms may have a practical application in developing tools for rapid prototyping of such systems.

In this chapter the idea of using optimal normal bases (o.n.b.) of second and third types in combination with polynomial basis of field

F

(

q

n

) is detailed using a new modification of o.n.b. called reduced optimal normal basis –1,

β

1

, …,

β

n

 − 1

corresponding to a permutated o.n.b.

β

1

, …,

β

n

 − 1

Operations of multiplication, rising to power

q

i

, rising to arbitrary power and inversion in reduced o.n.b. in combination with polynomial basis as well as converting operations between these bases in the fields of characteristic three has been described, estimated and expanded to the fields of characteristic two. This allows get efficient implementations of cryptographic protocols using operation of Tate pairing on supersingular elliptic curve.

Domain model is the first model in the scope of interests of software developers. Its creation, especially for complex domain, can be very costly and time consuming, as it extensively involves domain experts. On the other side, domain knowledge could be included in existing ontologies, and can be extracted, with the support of domain experts, from them. That way of knowledge extraction should lead to better model quality, preventing from misunderstandings between business analysts and domain experts. The paper presents an approach to business model development on the base of SUMO ontologies. The approach is explained with the use of simple, but real example. The results are promising. Domain models, created using this approach, could be perceived as valuable input for further development.

Business rules are an important part of requirement specification, and an essential input for software analysis and design. Usually, at the beginning, they are expressed in natural language, which is later translated by a business analyst to a more formal representation, e.g. UML diagrams. The translation process is error prone because business analysts can misinterpret or omit informally expressed business rules. The aim of the paper is to present an approach to automatic verification of UML class diagrams against business rules, expressed in a semi natural language, i.e. SBVRSE. The proposed approach has been implemented as a tool, and tested on representative examples. At that moment it supports structural business rules. In the future the method will be extended to cover also other types of business rules.

The problem of measuring vehicle’s weight-in-motion (WIM) is one of the most important research topics in the field of transport telematics. It is important not only for development of intelligent systems used for planning and cargo fleet managing, but also for control of the legal use of transport infrastructure, for road surface protection from early destruction and for safety support on roads. Data protection plays one of the crucial role in such kind of systems as data transmitted over internet network can be not only intercepted and disclosed, but also rigged and be used as a tool for attack on your equipment or system. Traditional data protection methods are increasingly becoming an easier barrier for implementing a successful hacker attack, but time for breaking existing encryption algorithms, which recently have had a high cryptographic strength, is gradually decreasing thus opening the door for developers to create new or upgrading existing encryption algorithms, whose characteristics will be able to withstand modern hacker’s threats.

The paper examines energy efficiency of running computational tasks in the hybrid clouds considering data privacy and security aspects. As a case study we examine CPU demanding high-tech methods of radiation therapy of cancer. We introduce mathematical models comparing energy efficiency of running our case study application in the public cloud, private data center and on a personal computer. These models employ Markovian chains and queuing theory together to estimate energy and performance attributes of different configurations of computational environment. Finally, a concept of a Hybrid Cloud which effectively distributes computational tasks among the public and secure private clouds depending on their security requirements is proposed in the paper.

The article presents a developed universal simulation model supporting the design process of the security control area at the airport. The universality of the simulation model allows for its use for the adaptation of the size of the security control area which consists of security check stations with a single flow of passenger streams to the forecast intensity of reporting passengers. The presented model is mostly intended for regional airports, where the security control is conducted using a metal detector and an x-ray device. The functional readiness of the designed system is analysed in terms of the forecast intensity of passenger reports. The functioning of the simulation model is based on time characteristics determined on the basis of research conducted on a real system which allowed for the verification of the functioning of the model. It is also possible to introduce one’s own characteristics to optimise another existing real system.

The article presents a developed simulation model of the check-in process at an airport. The developed algorithm of the process allows for an analysis of the entire system for any preset input parameters. The effectiveness of the functioning of check-in desks is measured by functional readiness, which can confirm both the reliability of the process in terms of the expected effectiveness and indicated the quality of the service offered (the time spent by the passenger in the system). The simulation model also allows for conducting an analysis of the possibility of controlling the input stream for security control, which is the next sub-process in the departing passenger check-in structure. Characteristics pertaining to the operation of the system were implemented in the model, such as the passenger service time and the structure of passenger reports to the system, which were determined on the basis of research conducted at a real airport.

The paper deals with the problem of analyzing performability of creating cloud environment within a server taking into account various hardware and software configurations. Within the hardware we consider the impact of using SATA and SSD discs for holding data and program codes of the involved processes. Within software we checked single and multiple thread approaches. The effectiveness of the activated processes was evaluated using performance monitoring based on vSphere tool. The performed measurements were targeted at CPU, RAM and disc usage as well as execution time analysis. We present the measurement methodology and interpretation of the obtained results, which proved their practical significance.

Searching in huge amount of information available on the internet is undoubtedly a challenging task. A lot of new web sites are created every day, containing not only text, but other types of resources: e.g. songs, movies or images. As a consequence, a simple search result list from search engines becomes insufficient. Recommender systems are the solution supporting users in finding items, which are interesting for them. These items may be information as well as products, in general. The main distinctive feature of recommender systems is taking into account personal needs and taste of users. Collaborative filtering approach is based on users’ interactions with the electronic system. Its main challenge is generating on-line recommendations in reasonable time coping with large size of data. Appropriate tool to support recommender systems in increasing time efficiency are clustering algorithms, which find similarities in off-line mode. Commonly, it involves decreasing of prediction accuracy of final recommendations. This article presents an approach based on clustered data, which prevents the negative consequences, keeping high time efficiency. An input data are clusters of similar items and searching the items for recommendation is limited to the elements from one cluster.

The trend of limiting vehicular traffic to the benefit of public transport is developed in contemporary urban planning. One of the tasks is determining location of the collective parking places in the P&R system. Criteria for assessing the quality of the selected location are formulated generally and descriptively. However, the factors to be assessed are often ambiguous and fuzzy, difficult to be precisely determined but possible for the evaluation by an expert. Due to the large number of parameters of criteria the practice has shown that the choice of the location of these sites in a way that is intuitive without a detailed analysis of all the circumstances, often gives negative results. Then the existing facilities are not used as expected. The authors have used fuzzy inference to the evaluation of the location of the P&Rs based on fuzzy input parameters. The obtained results of the analysis allows to determine the degree of attractiveness of the selected place on the basis of a broad set of the expert input data. The proposed evaluation method has been tested on three existing facilities for which the effect is already known.

In this paper the neural network controller for quadrotor steering and stabilizing under the task of flight on path has been deliberated. The control system was divided into four subsystems. Each of them is responsible for setting the control values for controlling position and speed of the quadrotor and for steering rotation speed of propellers. The neural network was taught by control system with standard PID controller. This approach is used for checking how neural networks cope with stabilisation of the quadrotor under flight task. Simulation results of the neural controller and PID controller working were compared to each other. The mathematical model of quadrotor and its neural controller were simulated using Matlab Simulink software. In the paper the simulation results of the quadrotor’s flight on path of are presented.

The process of ensuring security of citizens requires access to information from sensors placed in various points of monitoring and data acquisition systems. IP video monitoring networks are nowadays the main element of a system that combines a variety of software and hardware architectures forming a heterogeneous environment. Ensuring implementation of monitoring services with a predictable quality level is an important challenge for monitoring systems. The paper includes presentation of a QoS platform that offers the possibility of support for monitoring services based on the information on video stream quality parameters. The proposed solution is part of the quality support system for monitoring and data acquisition system implemented in the INSIGMA project.

Traditionally, the fault tolerance is a matter of redundancy, where hardware components are replicated or additional considerable lines of program code are inserted. In spite of the widespread usage even for the basic voted-groups architectures, the problem of complexity always has to be taken into account and could lead to decrease system reliability. In this paper the combined software and hardware methods to achieve necessary system requirements without enlarged implementation price and complexity is proposed. The reducing of system complexity helps to make up to 50% savings in development life cycle stage with higher availability and higher safety.

Digital networked systems become increasingly important and perform indispensable function in safety-critical systems. However, they are exposed to various networked degradations that affect their reliability which has not been widely studied. In this paper, the reliability of such systems is estimated using event-based Montel Carlo simulation assuming a time-varying model. The degradations are described by Markov process and multi-state Markov chain subject to uncertainties. A case study is provided to illustrate the efficiency of the proposed framework.

In the presented paper, the authors’ research work is focused on the analysis of maintenance decision making process performance with taking into account necessary operational data availability. Thus, in the Introduction section, the transportation systems maintenance issues are described. Later, there is a comprehensive literature overview in the analysed research area provided. This gives the possibility to present the decision making process in the transportation systems’ maintenance management area. Later, in the next Section, the case study for maintenance management processes performance in chosen passenger transportation company is investigated. Following this, the computer systems used for operational data gathering are characterised, and the data availability is investigated.

The effective reuse of design assets in safety-critical Software Product Lines (SPL) would require the reuse of safety analyses of those assets in the variant contexts of certification of products derived from the SPL. This in turn requires the traceability of SPL variation across design, including variation in safety analysis and safety cases. In this paper, we propose a method and tool to support the automatic generation of modular SPL safety case architectures from the information provided by SPL feature modeling and model-based safety analysis. The Goal Structuring Notation (GSN) safety case modeling notation and its modular extensions supported by the D-Case Editor were used to implement the method in an automated tool support. The tool was used to generate a modular safety case for an automotive Hybrid Braking System SPL.

This paper is devoted to discussion of econometric techniques, utilized for analysis of bus stop spatial allocation. Majority of researches are focused on bus route characteristics, averaged by route (average distance between bus stops) or by time (daily ridership on a bus stop level). Meanwhile, modern electronic ticketing systems are widely used and their detailed data can be used for analysis with advanced econometric techniques. We discussed shortcomings of existing empirical researches in this area and proposed a list of econometric models, which can be applied for deeper analysis of bus stops. We suggest possible specifications of spatial regression models, which take spatial dependence between individual bus stops and spatial heterogeneity into account. Also we discussed possible applications of the stochastic frontier model to bus stop benchmarking.

Communication processes have to be observed because there are possibilities that a different kind of threats will occur in the processes of exchanging information in a network. These threats are connected with: the possibility of decryption, losing jurisdiction, believing in and freshness of information, message interception by intruders, etc. We also consider the run of the communication protocol operation. Security attributes have been introduced to analyze the chosen aspects of security, which are proposed by Burrows, Abadi, Needham [4] and others. They have created the system of rules that defines interrelated parts of communication operations with security aspects. In this research we continued the analysis of security in the direction of building the model of auditing and dynamic modification of chosen factors (adequate to the security aspect) with the possibility to form a prognosis.

The presented method for the assessment of water supply system functioning in face of undesirable events occurrence takes into account the failure analysis. The developed analysis will allow decision support in the economic efficiency assessment of the water supply infrastructure functioning. The proportional hazard method for the failure of water pipes assessment enables to assess the reliability of water pipes and to obtain the possibility of prediction of undesirable events. The use of the proposed method can significantly support management analysis of water supply system. The establishment of the assessment parameters is very important issue that requires the use of waterworks experience. Moreover, it is based on the real data from the water supply system functioning.

The paper focuses on problem of credential revocation in distributed environment where credentials are dispersed among users and there is no simple way to cancel already issued credential. To overcome presented problem, access requests acceptors must contact each credential issuer in order to check its validity status. This process is not always acceptable as it imposes more system load. The work focuses on RT Framework and RT credentials. The proposed solution associates with each credential a freshness constraint which defines how long the credential can be regarded as valid after last validity check. Constraints are propagated along a credential chain that makes the user be granted access to a shared resource. Article presents model and its semantics. The semantics associates each RT credential with a freshness constraint that should be used during credential validity check. The solution provides mechanism for cancelling RT credentials that allows controlling access to shared resources in a fine manner and limit system load.

An intensive growth of population, progressing processes of social and economic development and increasing urbanisation of fast-developing countries considerably contribute to the increase in the amount of the waste produced. This necessitates reasonable decisions to be made within waste man-agement in urban areas. Therefore, a clear increase in the interest in matters as-sociated with the selection of relevant strategies for planning the waste man-agement process has been noticeable in recent years. One of the global aims is to develop appropriate waste recycling strategies. Solving problems associated with the selection of an appropriate strategy for waste management, planning collection processes, sorting and recovery, designing infrastructures for waste treatment and monitoring of such processes often requires that a model-based approach should be applied. The main objective of this paper is to present a simulation model that aids the decision-making process as regards the planning of sorting processes for the selectively collected municipal waste at the Recy-cling Centre in WPO ALBA S.A.

Newest technologies like cloud computing are readily explored by both private users and organizations. Cloud brings with it ease of use and convenience for the end user. However often at the cost of convenience we may pay in terms of data security.

The paper presents a new access control security model that can strengthen data protection in public software as a service cloud. The new model gives the new opportunities and possibilities for tighter protection of data. Its constant checking of the permissions, dynamics, manageability and elasticity show that it can be introduced to the cloud environment and ensures tight security.

Assessment and assurance of conformity with regulation documents assumes significant cost in modern economies. Demonstration of compliance with security standards involves providing evidence that the standards’ security criteria are met in full substantiating appropriate decision. Nevertheless despite its importance such type of activity haven’t been addressed adequately by the available solutions and the tool support given to conformity assessment and assurance processes is rather poor. International standards do not contain any formal technique for security evaluation, what makes performing evaluation process complicated and one-sided. In the article the approach to the security assurance evaluation Advanced Security Assurance Case (ASAC) is proposed based on refined definition of existed assurance case structure.

The aim of this article is to present drawbacks of the most popular version/ application of virtual private network service based on Internet Protocol Security (IPsec) as well as to describe the most interesting alternatives used to develop modern business services. Firstly, the article presents history of virtual private network (VPN) and focuses particularly on Secure VPN, where data are encrypted. Secondly, it discusses various aspects of using IPsec VPN while requirements of enterprises are constantly rising. Thirdly, it indicates examples of DMVPN (Dynamic Multipoint VPN) and GETVPN (Group Encrypted Transport VPN) used for implementing private services in hub-and-spoke and full-mesh architecture. DMVPN is generally recommended for usage over public networks, where it creates a VPN and secures it. DMVPN is a very good security improvement for MPLS VPN. GETVPN is favorable to secure existing VPN over private networks. Contrary to DMVPN, it uses distinct polices and multiple overlays, which give limited interoperability.

In this paper, the author’s research work is focused on state defining method for modeling of complex transportation systems reliability and safety, especially the railway transportation system. The paper begins with an introduction related to a literature review on railway transportation system functionality, reliability and safety modeling. The set of states can be divided into two subsets: availability set and failure set of states. Defining states in terms of railway transportation system is a complex issue, therefore the state classification into availability an failure is not sufficient. In addition to the technical effects associated with incorrect operation of system components, traffic consequences (traffic disruptions) of events are important, especially in case of railway transportation system. The paper ends with conclusions of the analysis, dealing with applicability of potential models to solve problems for the real system and a summary with prospects for further research.

This paper is result of technical feasibility study done in Transport and Telecommunication Institute, located in Riga, Latvia. As many world cities, Riga as capital of Latvia is suffering from transport congestions. There are many objective reasons of this situation, like outdated transport infrastructure, star-shaped roads schema, limited number of possibilities to cross the river etc. Riga City Traffic Department has a number of solutions, but most promising is the idea of implementation of ITS. The ITS provides a number of solutions, but the target of completed technical feasibility study is related with implementation of on-line traffic forecasting tool, which is based on macroscopic transport model and could be treated as subsystem of the traffic management system. The results of analysis give a number of “hot spots” which should be taken into account before technical implementation of the system. But the analysis results lead to the conclusion that the implementation of the on-line traffic forecasting tool is feasible for Riga city.

In Poland, a problem of lack of radar coverage at certain flight levels over certain regions of the country is being observed. Therefore, the Polish Air Navigation Services Agency is looking for some better surveillance solutions for airspace and airport surface.The paper presents an analysis of surveillance systems with particular emphasis on the of the possibility of integrating radar systems with multilateration and automatic dependent ones. Presented article turned out to be basis for discussion on sense of implementation the integrated system of surface surveillance at the airport where such a surveillance is carried out only visually and there is a bigger risk of causing a dangerous situation.

The paper presents a formal semantics of decision-making based on Dempster-Shafer belief structures. We introduced a method of decision support taking into account the subjective expert information formalized in the form of family of estimations based on the combination of hypotheses and ordered weighted average operators. The task is formulated in terms of the belief structures and allows evaluating the minimum and maximum objectives through different types of aggregation operators. In the context of critical infrastructure management, our research shows that the matrix of possible solutions can be represented as a payoff matrix including performance indicators or in the form of a risk matrix corresponding losses on the specific combinations of decisions. To ensure variation in the goals we use different types of ordering alternatives depending on the type of the specific problem. Finally, an illustrative example was given to selecting strategies and prioritizing decisions to mitigate targeted cyber intrusions fit to effective IT-security risk management in different critical application.

With the development of the Distributed Generation (DG), the role of power distribution networks becomes more and more important, and their configuration tends to change from radial to meshed. In this paper, algorithms for planning meshed networks are proposed. The problem of network planning has been transformed to an operational research problem. The algorithms are automated to reduce the planning time. Additional checks have been added to reduce the total length of the final network. A reliability assessment has been proposed to locate the weak spots of a network and to compare different solutions. Finally an application of the proposed algorithms is demonstrated.

Due to the steadily increasing air traffic, the problem of selection of the appropriate landing of the aircraft in the absence of the possibility of a touchdown at the destination airport is becoming increasingly important. The article presents the results of the analysis of elements affecting safety and traffic flow during the operation of an aircraft emergency landing at the airport. The article presents the essential elements of airport infrastructure affecting the safe touchdown the aircraft, as well as presents the basic elements that affect the occurrence of an emergency landing.

This paper presents a method to identify the operational state of the data communications system (ICT). The article presents the support of the management and operation of process simulation. The presented method based on rough set theory. Presented calculation of coefficient supporting decision about operational state of system ICT. At the end of the article presents the results of a simulation program written by the author. Computer simulations have shown the scope of the analysis method and scope of coefficient supporting decision.

This work evaluates implementation efficiency of different cryptographic algorithms in selected hardware organizations and in different FPGA devices. The tests included AES symmetric cipher and two more contemporary hash algorithms: Salsa20 and Keccak-

f

[400] permutation function. Each algorithm was realized in hardware in five organizations: the basic iterative one, two with the loop unrolled and two with the loop unrolled and pipelined, then automatically implemented in two popular-grade FPGA devices from Xilinx: Spartan-3 and Spartan-6. Results of 30 test cases allowed for evaluation of particular strengths and weaknesses of the ciphers, the organizations and the FPGA architectures. In particular, the evaluation took into account implementation efficiency offered by the two device families, scalability of the ciphers with the loop unrolling factor and specific routing problems which came out in some configurations.

Matlab/Simulink is a widely used modeling notation for control systems design in automotive industries. Safety standards, such as ISO 26262, are emphasizing model-based testing, in which, test cases derived from the design model are used to show model-code conformance. In this paper, we propose a new aspect-oriented test case generation approach called “MB-ATG” from Simulink models. This approach exploits model checking technique capability to generate counterexamples that constitute test cases. We experiment a real automotive Simulink model with MB-ATG prototype to show its performance. Experimental results show that MB-ATG approach is compliant with standard structural coverage criteria and does not provide redundant test cases.

In this work, we describe approaches to creation of a ranked jobs framework within the model of cycle scheduling in Grid virtual organizations with such quality of service (QoS) indicators as an average job execution time and a number of required scheduling cycles. Two methods for job selection and scheduling are proposed and compared: the first one is based on the knapsack problem solution, while the second one introduces a heuristic parameter of job and computational resources “compatibility”. Along with these methods we present experimental results demonstrating the efficiency of proposed approaches and compare them with random job selection.

The paper presents a service oriented, online engine for processing and clustering texts in the Polish language. The engine, designed according to Web-Oriented Architecture paradigm, allows to run a large number of different language tools (like tagger, named entity recognizer, feature extractor) and clustering tools (like CLUTO or R) from almost any type of applications including HTML/JavaScript’s ones. It allows constructing of a complex workflow, not only a simple chain of tools. To meet high availability requirements, the engine is deployed in a private cloud.

Along with the dynamic evolution of wide area networks and network technology development, data security is becoming increasingly important. This issue is particularly important in organizations or companies with branch offices in many distant places all over the world where communication is required in real-time often and data leakage may even cause the collapse of the company. In order to guarantee the security of transmitted data between remote sites such as the central division office of the branches and mobile workers is proposed to apply technologies that use secure, encrypted tunnels such as virtual private network. The latest solution is a dynamic multipoint virtual private network technique that eliminates defects of previous versions. This paper verifies the efficiency of data privacy and protection afforded by IT.

Presented are new methods of loop execution time estimation for parallelized and distributed systems. The proposed solutions take account of data transfer time, data locality and synchronization of threads. In this way the methods have been adjusted to modern parallel and distributed systems, which permits to estimate execution times of loops compatible with FAN, PAR and PIPE transformations. The model-based estimates have been compared to real measurements of program loops parallelized in the OpenMP standard and adapted to distributed systems satisfying the MPI standard. The presented approach can be used for optimized allocation of tasks in multithreaded processors and in distributed systems.

The concept of information mobility and its application spheres have been overviewed in the article. The main task in the way of infomobility implementation is to create and manage the information systems that facilitate the use of the transport system by its potential users. The information systems for public transport in Latvia have been analysed on the information services offered. Based on the received results of the development, the recommendations of the infomobility service level improvement in Latvia have been formulated.

A systematic approach to urban transport system planning and managing means the inclusion of a systematic monitoring system to collect the necessary data and periodically updating the DSS databases, as well as updating of models in their repositories. This should be supported by introduction of new data and information without changing and deleting the old. The authors proposed the application of regression analysis for data actualisation and new obtaining, and considered several task settings for realization of such approach. The proposed methodology focuses on the issues of data updating and preparation for modelling, consideration of model preparation and simulation scenarios including the analysis of the influence of the new solution implementation on the neighbouring fragments of the network. The approach has been approved using the simulation model for a fragment of Riga City. The offered procedures can be used in the frame of model-driven DSS and give the possibility to fulfill the process of the model actualisation faster and less expensive without loss of accuracy.

A lot of mathematical approaches are used in importance analysis, which permits to investigate influence of system component state changes on the system reliability or availability. One of these approaches is Logical Differential Calculus, in particular Direct Partial Boolean Derivatives. A new algorithm for the calculation of Importance Measures with application of Direct Partial Boolean Derivatives is proposed in this paper. This algorithm is developed based on parallel procedures.

The article contains a method for distributing containers on the yard while unloading a container train. The method assumes a definition of the sequence of operations and their adjustment to the time of the container stay on the yard. As compared to the traditional approach to the issue, the proposed method allows for saving the load service time and further savings connected with the machine service cost.

One of the most commonly applied optimizations to Byzantine fault-tolerant replication is batching. Such approach involves packing multiple client requests into a single instance, thus reducing the per-request overhead and providing a potential increase in throughput. Existing solutions use either constant-sized batches or determine their sizes based on the performance of the underlying replication protocol. In this article we propose a different approach and introduce a method for selecting batch sizes that minimize the cost of cryptographic operations performed by replication protocols. The results of performed experiments show that our method can obtain up to 50% increase in throughput when compared to existing batching schemes. The proposed approach can be applied not only in modern BFT replication protocols, but also in solutions using other fault models, as long as they use some form of message authentication.

In timed Petri nets temporal properties are associated with transitions as transition firing times (or occurrence times). Specific properties of timed nets, such as boundedness or absence of deadlocks, can depend upon temporal properties and sometimes even a small change of these properties has a significant effect on the net’s behavior (e.g., a bounded net becomes unbounded or vice versa). The objective of symbolic analysis of timed nets is to provide information about the net’s behavior which is independent of specific temporal properties, i.e., which describes preperties of the whole class of timed nets with the same structure.