Search

Assistive technology promotes greater independence for people with disabilities by enabling them to perform everyday tasks while providing improvements or changing the methods of interacting with the technology required to perform such tasks. The cognitive and physical characteristics of people with Down syndrome are usually characterized by some form of difficulty in communication and learning and unique physical characteristics, which makes performing daily activities challenging and often require the help of an assistant. For this paper, the authors conducted a survey, and the collected data served as a basis for understanding the development of people with Down syndrome, movement in traffic and the environment, and the frequency and knowledge of information and communication technologies. The paper aims to propose CAT assistive technology model and conceptual system architecture for delivering information services based on the Internet of Things concept to increase the independence and mobility of people with Down syndrome.

For postal service providers to meet the growing expectations of users, technological processes organization is one of the most important factors. In the context of the Industry 4.0 and the overall digitalization of the business processes, the need for the postal service providers to adapt to new challenges arises. Postal service provider face challenges, not only to transform their technological processes, but to also adapt to competitive market and growing consumer needs and expectations. This research focuses on the proposal of improvement of postal service’s business processes for several types of problems in the context of the digital transformation and the Industry 4.0.

A technologically based society geared towards every human being represents a Society 5.0 environment whose goal is to upgrade society to improve daily human life. It can be achieved through the significant application of information and communication networks, technologies, services, and user equipment. Industry 4.0 is the basis for developing the Society 5.0 environment in which the harmonious coexistence of people and machines comes to the fore. Artificial intelligence is one of the critical technologies for the further growth and development of the Society 5.0 environment, whose ability is to convert collected networked stakeholder's data into values. The authors in this paper systematically review the literature to investigate the application of artificial intelligence technology within the Society 5.0 environment. This paper aims to present the possibilities and scenarios of using artificial intelligence in different areas of the Society 5.0 environment, such as the digital transport system, healthcare, and manufacturing.

Nova generacija optičkog umrežavanja javila se kao rezultat povećanih korisničkih zahteva i potrebe za efikasnim i ekonomičnim korišćenjem spektra i raspoloživog propusnog opsega. Eksponencijalno povećanje obima saobraćaja u optičkim transportnim mrežama neminovno dovodi do dostizanja granice fizičkog kapaciteta optičkog vlakna, tako da će u bliskoj budućnosti zahtev za racionalnim korišćenjem ovog resursa predstavljati imperativ. Izvesno je da trenutno zastupljena tehnologija multipleksiranja po talasnim dužinama neće biti u mogućnosti da podrži buduće saobraćajne zahteve. Fiksni razmak između kanala za posledicu ima veliki deo spektra koji nije iskorišćen, a posebno kod prenosa velikim brzinama. S tim u vezi, aktuelna su rešavanja određenih pitanja koja se odnose na problematiku rutiranja i racionalnog korišćenja optičkih resursa, jer se time može omogućiti dodatni protok na postojećoj infrastrukturi. Optički link prate i visoke cene pa je cilj operatora da sve resurse u mreži implementira i iskoristi na što racionalniji način što je osnovni motiv za istraživanja sprovedena u okviru doktorske disertacije. Istraživanja su usmerena na odgovarajuće tehnike saobraćajnog inženjeringa kojima se postiže preraspodela saobraćajnih tokova u mreži i racionalna upotreba resursa. Time se stvaraju uslovi za opsluživanje dodatnih saobraćajnih zahteva. U istraživanju su ispitani efekti primene konverzije talasnih dužina i parametri koji najviše utiču na performanse mreže. Lokacije i broj konvertora su istražene za različite topologije i stepen povezanosti mreža, intenzitet saobraćaja i maksimalan broj talasnih dužina. Agregacija saobraćaja, koja podrazumeva izvorno grupisanje konekcija, posebno je istražena u elastičnim optičkim mrežama koje se mogu posmatrati kao tip WDM (Wavelength Division Multiplexing) mreža u kojima se kao jedinični resurs posmatra frekvencijski slot umesto talasne dužine. Takve mreže omogućavaju fleksibilno podržavanje različitih binarnih protoka prema specifičnim zahtevima korisnika. Optički predajnik generiše izvorno grupisane konekcije čime je moguće smanjiti broj zaštitnih opsega i ostvariti uštede optičkog spektra a istovremeno smanjiti potreban broj optičkih predajnika. Nekoliko grupisanih konekcija naziva se optički tunel. Za rešavanje navedenog složenog optimizacionog problema predložen je algoritam zasnovan na metaheuristici Optimizacija kolonijom pčela. Rezultati pokazuju da se primenom predloženog algoritma mogu postići značajne uštede optičkih resursa, a time i unapređenje performansi mreže. S obzirom da su istraživani statički saobraćajni scenariji, predloženi algoritam i tehnike saobraćajnog inženjeringa mogu se efikasno primenjivati prilikom planiranja i dizajniranja optičkih mreža. The new generation of optical networking raised as a result of increasing users' traffic demands and the need for efficient and cost-effective spectrum and bandwidth usage in. Exponential growth of traffic volume in optical transport networks inevitably results in reaching the physical fiber limit capacity and hence, its rationale usage will be an imperative in the near future. It is certain that currently deployed wavelength division multiplexing technology will not be able to cope with the future traffic demands. Fixed channel spacing results in a large spectrum wastage, especially for high speed transmissions. Therefore, the issues related to the routing problem and rationale spectrum usage are of the interest because the additional capacity can be provided at the existing infrastructure. The optical link is followed by high prices and the operator’s goal is to implement and use all the network resources in the most rational manner possible, what is the main motivation for the research conducted within this dissertation. The research is focused on appropriate traffic engineering techniques that achieve the traffic flow redistribution in the network and rational use of resources. That makes conditions for serving additional traffic demands. The effects of applying wavelength conversion and parameters which most influence the network performances are investigated. Locations and the number of converters are investigated under different network topologies and connectivities, traffic load and maximum number of available wavelengths. Traffic grooming, that means source connections grooming has been particularly investigated in elastic optical networks, which represent a form of WDM (Wavelength Division Multiplexing) networks where a frequency slot is treated as a unit resource instead of wavelength. Such networks have the ability to adjust data rates to fit specific user demands. Groomed connections are generated in optical transmitter, thus making it possible to eliminate the necessary guard bands and in such way to save optical spectrum, while reducing the required number of transmitters. Such group of groomed connections is called an optical tunnel. To solve this complex optimization problem, a new algorithm based on the Bee colony optimization metaheuristic has been proposed in this thesis. The obtained results show that its implementation leads to significant optical resource savings and consequently better network performances. Since static traffic scenario has been considered, the proposed algorithm and traffic engineering techniques can be efficiently used for planning and designing of optical networks.

Current diameters of digital transformations within the Industry 4.0 paradigm identify significant challenges in enabling individual interactions with end-users depending on user requirements. For this reason, concept of Industry 5.0 appears. It implies the relevant stakeholders (employees, users, and devices) mutual interaction to address the need for personalization and mass customization of products for end-users. Society 5.0, as the latest paradigm, focuses on placing humans at the center of technological transformation and industrial automation driven by Industry 4.0. This paper aims to use analysis and synthesis methods of currently available scientific and professional literature to explore the possibilities provided by the intensive application of information and communication networks and technologies for establishing the Society 5.0 environment. Given that there are not many scientific and professional papers in this field, this paper's results will serve as a basis for understanding this paradigm and possible impact on production processes in the future.

New sorting center (NSC) of Croatian Post (HP) was build and opened in 2019. Within the NSC, advanced technologies enable more efficient processing and sorting of shipments and increase postal service quality. Six modern machines improve logistics and postal services’ performance and facilitate postal and logistics center processes. Also, the schedule of devices, departments, premises, and places of receipt and dispatch are optimally located. Shipments are received as soon as possible, sent for processing, and prepared for dispatch. This paper analyses the technological and technical indicators of the NSC.

Smartphone data traffic follows the rapid growth trend which is primarily driven by the simultaneous increase in the number of smartphones and the increasing amounts of data traffic generated by each device. Video-generated traffic is the dominant form of data traffic, and predictions are that in the next few years more than 80% of data traffic will be related to the video. On the other hand, the dominant segment of communication is achieved through instant messaging applications such as WhatsApp, Viber and Messenger. They are also increasingly used to transmit videos in communication between users. This paper analyzes the amounts of data traffic generated by popular instant messaging applications when sharing a one-minute video. Based on the results of various applications for measuring the amount of generated data traffic, similarities/differences in data traffic generation of instant messaging applications will be identified.

Devices that coexist in the Internet of Things environment are growing in number and their application is becoming more diverse. This opens many problem areas for research. Examples of such areas are the classification of IoT devices, the detection of network traffic anomalies that such devices generate, and the monitoring and management of IoT devices and communication infrastructure. Certain researches indicate homogeneity of device behavior within individual groups of IoT devices and heterogeneity between different groups. The problem of defining groups of devices with similar characteristics is emphasized. This paper presents an analysis of network traffic features generated by IoT devices in order to gain insight into the traffic features that can be used as a framework for further research in a field of device class definition and device classification in the IoT environment.

Nove širokopojasne tehnologije i stalno rastući obim saobraćaja imaju veliki uticaj na upravljanje infrastrukturnim kapacitetima telekomunikacione mreže. Efikasno planiranje mrežnih resursa nije moguće realizovati bez primene pouzdanih modela za prognoziranje tražnje, kao i precizno odabranih veličina za karakterizaciju saobraćaja. U tom smislu, osnovni predmet istraživanja ove doktorske disertacije su strategije proširenja mrežnih resursa, koje se oslanjaju na primenu različitih metoda planiranja telekomunikacionih mreža, sa ciljem određivanja optimalnog vremena za realizaciju proširenja. Predloženi modeli proširenja resursa mreže baziraju se na optimizaciji perioda proširenja resursa mreže uz minimizaciju troškova zagušenja i troškova neophodne opreme, pri čemu se ulazni podaci oslanjaju na prognozirane vrednosti tražnje na telekomunikacionom tržištu. Mogućnosti primene različitih metoda planiranja, u kontekstu strategije proširenja mrežnih resursa, razmatrane su na hipotetičkoj WDM mreži, sa koegzistencijom fiksnih i fleksibilnih grid tehnologija. U disertaciji su razmatrani suprotstavljeni zahtevi investicionih ulaganja u cilju brzog povraćaja uloženih sredstava i kvaliteta opsluživanja korisnika. Kompromisno rešenje predstavlja i novi pristup, predložen u disertaciji, koji se odnosi na izbor parametara za karakterizaciju mrežnog saobraćaja, sa ciljem smanjenja uticaja neizvesnosti prognoziranih podataka na odluku o proširenju kapaciteta. U tom smislu, izvršena je analiza, na osnovu koje je utvrđeno da se, poređenjem vrednosti nivoa penala funkcije troškova zagušenja, sa jedne strane i vrednosti parametra Verovatnoće blokiranja kapaciteta, sa druge, može odrediti optimalni vremenski period za sprovođenje odluke o migraciji tehnologija, u posmatranom slučaju. Među najvažnijim naučnim doprinosima disertacije, mogu se izdvojiti implementacija prognoziranih vrednosti saobraćajnih zahteva, zasnovana na primeni teorije difuzionih modela, novi optimizacioni model, kao modifikacija modela strategije serijskog proširenja kapaciteta mrežnog linka, u kojem se uvodi mogućnost prelaska na novu tehnologiju, kao i novi pristup za smanjenje uticaja neizvesnosti prognoziranih podataka na određenu odluku o proširenju kapaciteta, zasnovan na izboru parametara za karakterizaciju mrežnog saobraćaja. New broadband services and ever-increasing Internet traffic volumes have a major influence on the backbone infrastructure capacity management, raising the importance of efficient planning of network resources, based on demand forecast and the appropriate selection of traffic characterization parameters, as well. Therefore, the main subject of the research, presented in the doctoral thesis are the network resources expansion strategies, with respect of different telecommunication network planning methods. Relaying on the forecasted demand in the telecommunications market, the proposed models are used in order to determine the optimal times for network capacity expansion and the technology migration, by minimizing the equipment and congestion costs. Application of different planning methods, in the context of the network resource expansion strategies, has been considered within the hypothetical WDM network, with the coexistence of fixed and flexible grid technologies. Having in mind the contrast between network performance improvement and the investors’ interest for the longer operation of already built-in technologies, a novel approach has been proposed in order to determine the appropriate time for making the technology migration. In this sense, the traffic characterization parameters (the Penalty function and the Bandwidth Blocking Ratio) were used. The main innovative aspect of the approach proposed in the thesis considers the combination of these two metrics in order to decrease uncertainty of the forecasted demands. The other important scientific contributions of the dissertation are the traffic demand forecasting approach based on the diffusion model theory, as well as the novel multiperiod network capacity expansion model, which introduces the option of the migration to the new technology.

Funkcija daljinskog upravljanja industrijskim sistemom postavlja specifične zahteve za informacionu i komunikacionu infrastrukturu, koja treba da obezbedi procesiranje i siguran prenos heterogenih informacija sa različitim zahtevima za kvalitet servisa. Komunikacija se ostvaruje između centra upravljanja i objekata industrijskog sistema, kao i između distribuiranih centara upravljanja. Usvajanje otvorenih komunikacionih standarda, korišćenje otvorenih softverskih platformi, povezanost sistema upravljanja sa drugim mrežama, daljinski pristup i dostupnost tehničkih informacija su razlozi zbog kojih je informaciona i komunikaciona infrastruktura savremenih industrijskih sistema daljinskog upravljnja, a posebno SCADA (Supervisory Control and Data Acquisition) sistema podložna različitim vrstama napada. Uzimajući u obzir evidentnu potrebu za implementacijom specifičnih mehanizama zaštite u mreži industrijskih sistema daljinskog upravljanja, poželjno je da se, pri projektovanju bezbednosnih sistema i kasnije u toku eksploatacije, izvrši procena bezbednosnog rizika, sa ciljem da se odredi racionalan nivo ulaganja u zaštitu. U disertaciji je prvo utvrđen stepen degradacije ključnih performansi telekomunikacione mreže SCADA sistema, simulacijom različitih uslova distribuiranih napada kao što je napad koji prouzrokuje odbijanje servisa (DDoS – Distributed Denial of Service). Zatim su predložena dva nova metoda procene bezbednosnog rizika u slučaju DDoS napada na infrastrukturu SCADA sistema. Prvi, osnovni metod, zasniva se na analizi arhivskih podataka, a pretpostavlja proračun povrata investicija u zaštitu pomoću skupa težinskih faktora, koji kvantifikuju uslove u kojima se dogodio napad. Drugi, hibridni metod, pored analize arhivskih podataka, uzima u obzir subjektivnu ocenu stručnjaka dobijenu na osnovu odgovarajućih anketa. U zavisnosti od primene metoda predložena su dva načina izražavanja mere rizika, kvalitativno i monetarno. Na kraju su predloženi postupci cost/benefit analize za preporučenu primenu IDPS (Intrusion Detection and Prevention System) mehanizama zaštite na osnovu procenjene mere rizika. Definisanje prihvatljivog praga za povrat investicija u zaštitu omogućuje donošenje odluke o racionalnom ulaganju u bezbednost SCADA sistema. Za testiranje predloženih metoda definisane su dve studije slučaja: studija slučaja u realnom okruženju protočne hidroelektrane i studija slučaja SCADA sistema u modelovanom magistralnom gasovodu. Rezultati studija slučaja su pokazali da su metodi pogodni za identifikaciju ranjivosti (vulnerability) sistema, praktični i primenljivi u različitim industrijskim sektorima. Pored toga, pokazalo se da su metodi efikasni u proceni mere bezbednosnog rizika od infrastukturnog napada i proceni isplativosti ulaganja u poboljšanje bezbednosti infrastrukture SCADA mreža. Studija slučaja u magistralnom gasovodu pokazala je da je drugi metod primenljiv i u fazi projektovanja sistema, kada arhive sa relevantnim podacima nisu dostupne. Na kraju disertacije, na bazi rezultata istraživanja, predložene su mere za ograničavanje bezbednosnog rizika u industrijskim sistemima daljinskog upravljanja. Remote control of industrial system poses specific requirements for information and communication infrastructure, which has to provide processing and secure transmission of heterogeneous information with different requirements for Quality of Service. Communication takes place between control center and industrial system devices, as well as among distributed control centers. Information and communication infrastructure of modern Supervisory Control and Data Acquisition (SCADA) systems is particularly vulnerable to different cyber security threats due to following reasons: adoption of open communication standards, use of open software platforms, connectivity with other networks, remote access, and availability of technical information. There is an evident need to implement specific security mechanisms in industrial control networks; hence, in order to determine a cost-effective level of investment, it is desirable to assess security risk during network design phase, as well as during network operation. In this thesis, we first investigate the level of network performance degradation in SCADA systems by simulation of different conditions of distributed attacks such as Distributed Denial of Service (DDoS). Further, two novel methods for security risk assessment are proposed for the case of DDoS attack on the SCADA system infrastructure. The first, basic method relies on the analysis of historical data, and assumes calculating return on security investment as a function of the set of weighting factors that quantify the attack conditions. The second, hybrid method takes into account both hictorical data and subjective assessment of experts, provided by appropriate questionnaires. Depending on method application two ways (qualitative and monetary) for expressing the risk measure are proposed. Finally, techniques of cost/benefit analysis are also proposed for recommended application of intrusion detection and prevention system, based on the assessed risk measure. Definition of acceptable threshold for return on security investment allows making decision about cost-effective level of investment in security of SCADA system. For testing of proposed risk assessment methods, two case studies are defined: the first one considers real environment of a run-off-river hydropower plant, and the second one investigates the SCADA system in a simulated main pipeline. The results of case studies have shown that proposed methods are suitable for identification of system’s vulnerability, useful and applicable in different industrial sectors. Besides, proposed methods are efficient in security risk assessment regarding infrastructure attacks as well as in analysis of investment feasibility regarding enhancement of the SCADA network infrastructure security. Case study of the main pipeline also shows that the second method is applicable in the system design phase when relevant historical data are not available.

With the development of assistive technologies in the past, people with disabilities have considerably increased the degree of mobility while doing their daily activities. According to the EU Council Directive 2000/78 / EU and the Employment Guidelines, every person with disabilities is important to provide employment opportunities and equal access to the wider community. Therefore, this paper will make an analysis of currently available research on the possibilities of applying Universal Design and forms of assistive technologies whose task is to enable greater involvement of people with disabilities in the field of employment. Based on the collected conclusions of the conducted analysis, the experience of users using assistive technology in their workplace will be presented. In addition to mentioned, the representation of information and communication technologies and its features will depend on the usage scenarios and the perspective of the future development of new services in the work processes of persons with disabilities and Industry 4.0.

Recent development and increase of e-commerce has brought new challenges to postal services market. The new requirements for the postal market should be considered Since e-commerce has started to grow. Focus should primarily be on the implementation of new technologies in business processes, with emphasis on the last mile delivery. This paper explores recent technological development that accomplish better information and commodities flow in a postal system to improve quality of service, as well as user satisfaction.

By using smartphones, users are generating data traffic. Users characteristics are an important factor in generating smartphone data traffic. This paper presents a methodology for measuring smartphone data traffic and collecting of data about the individual characteristics of users of those smartphones. Measurement of generated data traffic is achieved by using an integrated smartphone application. The individual characteristics of smartphone users are determined on answers collected through the online questionnaire. The results of the developed methodology enable anonymous collection of aggregated smartphone data traffic and the characteristics of users of those smartphones.

Predmet istraživanja ove doktorske disertacije je аnаlizа difuzionih modelа nа osnovu koje je predložen i razvijen novi model prognozirаnjа novih servisа sа promenljivim potencijаlom tržištа. Novi model se bazira nа modifikаciji osnovnog Bass-ovog difuzionog modelа u cilju prilаgođenja modelа eksploаtаcionim kаrаkteristikаmа servisа i razvoju neophodne infrastrukture mreže, kao i socio-ekonomskim kаrаkteristikаmа potencijalnog tržišta. Pored rаzvojа novog modelа, dаt je i prikаz postojećih modelа prognozirаnjа dostupnih u stručnoj literaturi. Osnovni cilj disertacije je rešavanje problema prognoziranja brzine difuzije prihvatanja novog komunikacionog servisa, koja zavisi od dinamike razvoja infrastrukture mreže. Ukoliko tehnička podrška novog servisa nije realizovana istovremeno, servis se ne može ponuditi kompletnom tržištu. S obzirom da se proces prihvatanja novog servisa odvija etapno, shodno izgradnji infrastrukture mreže, potencijal tržišta se ne može tretirati kao vremenski nezavisan parametar modela (osnovni nedostatak Bass-ovog modela). Upravo to je i osnovni motiv disertacije – razviti difuzioni model koji potencijal tržišta posmatra kao vremenski zavisnu veličinu, čiju promenu diktira dinamika razvoja infrastrukture komunikacione mreže. Nakon izučavanja različitih difuzionih modela, tehničko-eksploatacionih zahteva novih servisa i socio-ekonomskih indikatora razvoja tržišta, predložen je novi model sa promenljivim potencijalom tržišta. U tezi je primenjena segmentacija tržišta prema socio-ekonomskim karakteristikama korisnika i topologiji terena, kao i simulacija različitih scenarija razvoja u cilju odabira optimalnog pristupa razvoju infrastrukture mreže. Dobijeni rezultati pokazuju da razvoj i implementacija novog modela mogu imati značajnu praktičnu primenu kod telekomunikacionih operatora i servis provajdera, u cilju uspešnog planiranja i projektovanja komunikacione mreže. Adekvatnim planiranjem mreže omogućeno je, kako poboljšanje funkcionisanja same mreže, tako i poboljšanje kvaliteta servisa koji se nudi korisnicima. Takođe, efekti koji se mogu prepoznati ogledaju se i u pravovremenom uvođenju novih servisa, koordinisanoj proizvodnji terminalne opreme, kao i u određivanju i rasporedu potrebne kadrovske strukture... The main research topic of this doctoral thesis is the analysis of new service diffusion forecasting models. Based on this analysis a new diffusion model with variable market potential is proposed. Development of new mathematical model is based on the Basic Bass diffusion model. The aim is to adjust the model to the operational features of services/products and to the socio-economic parameters of the market that drive potential customers. In addition, this thesis considers existing forecasting models available at the literature and the obtained results of the proposed model are compared to the similar forecasting models. This thesis tries to solve the problem of forecasting the adoption of new service that depends on the dynamics of network infrastructure development. If technical support of some new service is not provided in the whole territory, then the service cannot be simultaneously offered to the market. Therefore, the adoption process is performing in steps, following the systematic process of the infrastructure development. The model considers variable market potential that depends on the technical capabilities related to some new telecommunications services. The availability of traffic area infrastructure is increasing during the time according to the infrastructure investment necessity. That is the basic motive of the thesis - developing the diffusion model with variable market potential, caused by dynamics of the network infrastructure development. After researching various diffusion models, as well as technical parameters of the services and socio-economic indicators of the market, an optimal forecasting model is proposed that includes the time varying market potential. This model uses market segmentation and the simulation of various market scenarios in order to indicate effective service rollout. Demand forecasting is required in order to dimensioning of required network resources. Moreover, this model offers the insight into the operator's compromises regarding decisions in the network investment. The resulting plan is a network deployment strategy that defines temporal sequence of network layouts. Thus, the upgrade plan indicates which investments are indispensible at each point of time. The model is of great importance for operators and service providers in decision making and planning, especially in developing countries, as well as in times of economic crisis. It should be noted that the strength of the model lies in its robustness and applicability to real situations that are especially characteristic for the telecommunications market...

The liberalization of the postal services market includes the obligation to implement the public procurement of postal services. This enables adequate market competition for bodies subjected to the Directive on public procurement. Because of mutually incomparable factors that affect the realization of services, proposed research presents a comprehensive analysis of the application of multi-criteria analysis and AHP method in public procurement, when selecting the most economically advantageous tender. The main problem that occurs is the choice of relevant criteria and related sub-criteria that have more or less impact on the postal services provider.

U ovoj disertaciji predmet istraživanja je rešavanje problema optimizacije resursa javnog poštanskog operatora (JPO). Predložen je model koji optimizuje ograničene resurse, troškove i operativne prihode korišćenjem neparametarske analize, DEA (engl. Data Envelopment Analysis). Testiranje i verifikacija predloženog modela izvršeni su na skupu od 27 JPO iz zemalja Evropske Unije i Srbije. Dobijeni rezultati eksplicitno pokazuju da su postojeći resursi i način kako se oni koriste ključni u postizanju rezultata rada, kao i to kako može da bude definisan efikasan operativni plan. Između ostalog, za slučaj javnog poštanskog operatora u Srbiji, rezultati ukazuju da ne postoji višak u broju zaposlenih, dok je višak prepoznat u ukupnom broju jedinica poštanske mreže. Pored toga, predstavljeno je izvođenje analize evaluacije efikasnosti javnog poštanskog operatora tokom vremena. U disertaciji, predložen je originalni višeulazni i višeizlazni model za merenje troškovne efikasnosti, CE (engl. Cost Efficiency) dostavnih jedinica poštanske mreže. Analizirana je dostavna mreža javnog poštanskog operatora u Srbiji koja uključuje 1,194 dostavne jedinice. Problem brojnosti i heterogenosti ovih jedinica prevaziđen je definisanjem originalnog algoritma za klasterovanje. Ovaj algoritam je obezbedio poređenje efikasnosti uporedivih jedinica, kao i troetapnu analizu: jedinice-klasterimreža. Dobijeni nalaz ukazuje da je posmatrana dostavna mreža troškovno neefikasna (CE=40%). Dalje, dekompozicijom CE objašnjeno je kako nivo troškovne efikasnosti može biti poboljšan. Ilustracije radi, utvrđeno je da najmanje 794 dostavne jedinice imaju mogućnost za poboljšanje samo na osnovu realokacije resursa. Istražen je uticaj ekonomije obima na efikasnost dostavnih jedinica gde je utvrđeno da ona ima jači uticaj na efikanost malih jedinica u poređenju sa velikim jedinicama. Konačno, na osnovu izvedenih rezultata, dato je obrazloženje zašto predloženi model za merenje troškovne efikasnosti može da bude dragoceno sredstvo za kreatore politike i regulatore iz poštanskog sektora, kao i za menadžere poštanskih operatora. Doktorska disertacija, Predrag V. Ralević vi U disertaciji je predstavljen novi koncept univerzalne poštanske usluge. Suština ovog koncepta je mogućnost da se pošiljalac opredeljuje za uslugu koja bi se obavila u roku do 24 časa od vremena predaje pošiljke na prenos ili za drugu vrstu prenosa koji bi podrazumevao duži rok. Na osnovu toga, predstavljene su tri alternative novog koncepta koje ukazuju da se redefinisanjem univerzalne poštanske usluge može doprineti boljoj diversifikaciji ponude poštanskih usluga. Sa druge strane, u slučaju JPO iz Srbije pokazano je da predloženi koncept nove usluge ima pozitivan uticaj sa stanovišta očekivanih prihoda. To potvrđuju rezultati koji su izvedeni korišćenjem analize vremenskih serija, regresione i korelacione analize, teorije difuzije za prognoziranje tražnje za novom poštanskom uslugom (Basov model), kao i metode upitnika i intervjua. In this dissertation the research subject was to solve the problem of resource optimization of Universal Service Provider (USP). The model optimizing the limited resources, expenditure and operating revenue by using nonparametric analysis (DEA) is proposed. The testing and verification of the proposed model were performed on a set of 27 USPs from European Union member states and Serbia. The obtained results explicitly show that existing resources and how they are used are crucial in achieving the performance outcomes as well as to define the efficient operational plan. Among other, in the case of the USP in Serbia, the results indicate that there is not an excess in the number of employees, while the excess is recognized in the total number of permanent post offices. Beside that, in this thesis the derivation analysis of efficiency change over time for universal service provider is presented. In the dissertation, a multiple-input and multiple-output model for measuring the cost efficiency (CE) of postal delivery branches is proposed. The complete postal network of the Serbian Post delivery offices, which includes 1,194 post office branches is analyzed. The problem of numerous and heterogeneous of delivery offices was overcome by defining a new clustering algorithm. This algorithm enable to compare the branches that operate under the same or similar conditions and a three-level analysis: branches-clusters-network. The findings obtained indicate that the observed delivery network is cost inefficient (CE=40%). Furthermore, the explanation of how the level of CE could be improved is given by the decomposition of cost efficiency. For illustration, the results of this thesis indicate that, at a minimum, 794 delivery branches could improve their efficiency by resource reallocation. An impact of the economies of scale to efficiency of delivery branches is investigated and it indicates that the economies of scale have a stronger impact on small branches compared to large units. Finally, based on obtained results, the rational of how suggested model for measuring the cost efficiency can be beneficial to policy makers from the postal sector and to company managers is provided. Doktorska disertacija, Predrag V. Ralević viii In the dissertation a new concept of universal service obligation is presented. The essence of this concept is a possibility that the sender choose a transit time for his/her shipment, should it be within 24h or more. Based on that, three alternatives to the new concept are presented and they indicate that can contribute to a better diversification of offer of postal services. On the other hand, in the case of the USP from Serbia is demonstrated that the proposed concept of new service has a positive impact from the standpoint of expected revenue. This confirms the results derived by using time series analysis, regression and correlation analysis, the diffusion theory for forecasting demand for new postal service (Bass Model) as well as interview and questionnaire methods.

Availability of information and services, along with integrity and confidentiality presents a critical parameter in security in information and communication systems. Activities focused on denial of network communication availability are current from the beginning of global communication network development and they demand continuous development of protection methods. Significant challenge is the emergence of the Internet of Things (IoT) concept which will significantly increase the number of connected devices. That kind of environment is possible to use for generation of DDoS attacks. The paper investigates the effect of a significant increase in the number of connected devices in the IoT concept on increase of the number and volume of DDoS attacks.