It is shown in the literature that the NAT devices have become a convenient way to hide the identity of malicious behaviors. In this thesis, the aim is to identify the presence of the NAT devices in the network traffic and (if possible) to predict the number of users behind those NAT devices. To this end, I utilize different approaches and evaluate the performance of these approaches under different network environments represented by the availability of different data fields. To achieve this, I propose a machine learning (ML) based approach to detect NAT devices. I evaluate my approach against different passive fingerprinting techniques representing the state-of-the-art in the literature and show that the performance of the proposed ML based approach is very promising even without using any payload (application layer) information.

FundaÃÃo Cearense de Apoio ao Desenvolvimento Cientifico e TecnolÃgico / Este trabalho apresenta uma proposta metodolÃgica alternativa para a modelagem macroscÃpica de trÃfego de veÃculos, considerando o prÃprio desenho da malha como indutor dos fluxos de veÃculos, constituindo-se num estudo de caso dos Volumes MÃdios DiÃrios da cidade de Fortaleza (CE). Foi utilizada a ferramenta chamada de Sintaxe Espacial e uma de suas derivaÃÃes, chamada de AnÃlise Angular de Segmentos, como critÃrio de hierarquizaÃÃo das vias. Os valores gerados por tal ferramenta foram utilizados como dados de entrada, juntamente com os dados operacionais de trÃfego para a construÃÃo de um esquema de treinamento para modelagem atravÃs de Redes Neurais Artificiais. Como dados de saÃda, foram utilizados os valores de VMD. O objetivo foi desenvolver um mÃtodo de baixo custo para estimar o volume trÃfego veicular e que eliminasse a necessidade de numerosos dados exigidos para a modelagem tradicional ou modelo âquatro etapasâ, como por exemplo a pesquisa origem / destino / This paper presents an alternative methodological proposal for macroscopic modeling of traffic flows rates, considering the grid design as inductor of veicular flows, configuring a case study of the Daily Average Volumes of Fortaleza City, Cearà State. The tool called Space Syntax and one of its derivations called Angle Segments Analysis were used as a criterion of hierarchization of the streets. The values generated by such tool were used as input data as well as operational data of traffic were used for the construction of a training scheme for modeling through Artificial Neural Networks, and the values of Daily Average Volumes were used as output ones. The research aims at developing low cost method to preview volume flows and that could eliminate the need of numerous data required by the traditional UTMS modeling know as of four-stage model like for example, the survey origin / destination

Traffic accidents is one of the major causes of fatalities and economic loss around the world. Thus, there is an urgent need for a better understanding about the factors that contribute to accidents so that the accidents can be prevented in the future. The research objective of this thesis is to analyze the traffic accidents between vehicles and unprotected road users (pedestrians and bicycles) in Stockholm, finding spatial distribution patterns, related attributes and examining relationships between accidents and a number of vehicle flows. The data is first analyzed with general statistical analysis to examine the basic characteristics. There is no apparent trend of change among the number of accidents per year, while the numbers of accidents happening from May to October is higher than the rest of the year except for July due to less traffic during holiday period. Most traffic accidents occur in overcast weather, on a dry road surface, or during the day. In the spatial analysis part of the thesis, Global Moran’s I is used to detect whether there is an attribute-related spatial distribution pattern. Hot spot analysis is then applied on the clustered attributes to find significant hot and cold spots over the study area. The conclusions are that road surface conditions and occurrence time during day/night are two related factors that influence traffic accidents while weather is not considered a related attribute since the accidents distribute randomly in terms of weather, of which it is difficult to obtain temporally-aligned, detailed local information for further analysis. Different parameters are selected and discussed during the process. When calculating the distance between two accidents in traffic accident analysis, Manhattan distance is more appropriate than Euclidean distance since traffic accidents are restricted to the road network. The distance band determines scales of analysis tools, with 50 meters on an intersection and 500 meters for a larger region in Stockholm. Most hot spots arise at intersections and roundabouts where different types of traffic flows meet each other. The result of the relationships between traffic accidents and different types of vehicle flows shows that the correlation coefficients between number of traffic accidents and traffic flows are low, meaning that there is no obvious correlation between them, which is also proved by the scatter plots. Poisson regression model is applied on the traffic accident data. As a result, high-risk and low-risk areas in Stockholm are pointed out. Some are consistent with the hot-spot analysis result.

Dynamic Traffic Assignment (DTA) is gaining wider acceptance among agencies and practitioners because it serves as a more realistic representation of real-world traffic phenomena than static traffic assignment. Many metropolitan planning organizations and transportation departments are beginning to utilize DTA to predict traffic flows within their networks when conducting traffic analysis or evaluating management measures. To analyze DTA-based optimization applications, it is critical to obtain the dual (or gradient) information as dual information can typically be employed as a search direction in algorithmic design. However, very limited number of approaches can be used to estimate network-wide dual information while maintaining the potential to scale. This dissertation investigates the theoretical/practical aspects of DTA-based dual approximation techniques and explores DTA applications in the context of various transportation models, such as transportation network design, off-line DTA capacity calibration and dynamic congestion pricing. Each of the later entities is formulated as bi-level programs. Transportation Network Design Problem (NDP) aims to determine the optimal network expansion policy under a given budget constraint. NDP is bi-level by nature and can be considered a static case of a Stackelberg game, in which transportation planners (leaders) attempt to optimize the overall transportation system while road users (followers) attempt to achieve their own maximal benefit. The first part of this dissertation attempts to study NDP by combining a decomposition-based algorithmic structure with dual variable approximation techniques derived from linear programming theory. One of the critical elements in considering any real-time traffic management strategy requires assessing network traffic dynamics. Traffic is inherently dynamic, since it features congestion patterns that evolve over time and queues that form and dissipate over a planning horizon. It is therefore imperative to calibrate the DTA model such that it can accurately reproduce field observations and avoid erroneous flow predictions when evaluating traffic management strategies. Satisfactory calibration of the DTA model is an onerous task due to the large number of variables that can be modified and the intensive computational resources required. In this dissertation, the off-line DTA capacity calibration problem is studied in an attempt to devise a systematic approach for effective model calibration. Congestion pricing has increasingly been seen as a powerful tool for both managing congestion and generating revenue for infrastructure maintenance and sustainable development. By carefully levying tolls on roadways, a more efficient and optimal network flow pattern can be generated. Furthermore, congestion pricing acts as an effective travel demand management strategy that reduces peak period vehicle trips by encouraging people to shift to more efficient modes such as transit. Recently, with the increase in the number of highway Build-Operate-Transfer (B-O-T) projects, tolling has been interpreted as an effective way to generate revenue to offset the construction and maintenance costs of infrastructure. To maximize the benefits of congestion pricing, a careful analysis based on dynamic traffic conditions has to be conducted before determining tolls, since sub-optimal tolls can significantly worsen the system performance. Combining a network-wide time-varying toll analysis together with an efficient solution-building approach will be one of the main contributions of this dissertation. The problems mentioned above are typically framed as bi-level programs, which pose considerable challenges in theory and as well as in application. Due to the non-convex solution space and inherent NP-complete complexity, a majority of recent research efforts have focused on tackling bi-level programs using meta-heuristics. These approaches allow for the efficient exploration of complex solution spaces and the identification of potential global optima. Accordingly, this dissertation also attempts to present and compare several meta-heuristics through extensive numerical experiments to determine the most effective and efficient meta-heuristic, as a means of better investigating realistic network scenarios. / text

The issue of transport is currently one of the most discussed topics, thanks to deepening regional differentiation causing increased demands on the quality of existing transport links. Especially the attractiveness of centers and the existence of good transport infrastructure, stimulating the development of the region (his success), is becoming one of the most debated issues of our times. This diploma thesis deals with the comparison and evaluation of changes in spatial interaction in the settlement of the Plzeň/ Pilsen region in two time slices (1991 and 2001). Analysis of the interactions of traffic will be done by using statistical methods and using GIS technology. Main attention will be paid to the spatial differentiation of the frequency of traffic flows and transport accessibility. Subsequently, the transport characteristics will be compared with the characteristics of the complex importance of settlement, so the basic relationships between these indicators can be found. Keywords: spatial interaction, accessibility, frequency of public transport, traffic flows, transport regionalization, Plzeň/ Pilsen region.

In this Internet and big data era, resource management has become a crucial task to ensure the quality of service for users in modern wireless networks. Accurate and rapid Internet traffic data is essential for many applications in computer networking to enable high networking performance. Such applications facilitate admission control, congestion control, anomaly detection, and bandwidth allocation. In radio networks, these mechanisms are typically handled by features such as Carrier Aggregation, Inter-Frequency Handover, and Predictive Scheduling. Since these mechanisms often take time and cost radio resources, it is desirable to only enable them for users expected to gain from them. The problem of network traffic flow prediction is forecasting aspects of an ongoing traffic flow to mobilize networking mechanisms that ensures both user experience quality and resource management. The expected size of an active traffic flow, its expected duration, and the anticipated amount of packets within the flow are some of the aspects. Additionally, forecasting individual packet sizes and arrival times can also be beneficial. The wide-spread availability of Internet flow data allows machine learning algorithms to learn the complex relationships in network traffic and form models capable of forecasting traffic flows. This study proposes a deep-learning-based flow prediction method, established using a residual neural network (ResNet) for regression. The proposed model architecture demonstrates the ability to accurately predict the packet count, size, and duration of flows using only the information available at the arrival of the first packet. Additionally, the proposed method manages to outperform traditional machine learning methods such as linear regression and decision trees, in addition to conventional deep neural networks. The results indicate that the proposed method is able to predict the general magnitude of flows with high accuracy, providing precise magnitude classifications. / I denna Internet och data era har resurshantering blivit allt mer avgörande för att säkerställa tjänstekvaliteten för användare i moderna trådlösa nätverk. Noggrann och hastig Internet-trafikinformation är avgörande för många applikationer inom datanätverk för att möjliggöra hög nätverksprestanda. Sådana applikationer underlättar kontroll av behörighet, kontroller av trängsel, detektering av avvikelser och allokering av bandbredd. I radionätverk hanteras dessa mekanismer vanligtvis av funktioner som Carrier Aggregation, Inter- Frequency Handover och Predictive Scheduling. Eftersom dessa funktioner ofta tar tid och kostar resurser så är det önskvärt att nätverk endast möjliggör sådana funktioner för användare som förväntas dra nytta av dem. Prediktering av trafikflöden i nätverk grundar sig i att förutsäga aspekter av ett pågående trafikflöde för att kunna mobilisera nätverksfunktioner som säkerställer både kvaliteten för användare samt resurshantering. Den förväntade storleken på ett aktivt trafikflöde, dess varaktighet och mängden paket inom flödet är några av dessa aspekter. Det kan dessutom vara fördelaktigt att förutsäga individuella paketstorlekar och ankomsttider. Den stora tillgången till data med nätverks-flöden gör det möjligt för maskininlärningsmetoder att lära sig de komplexa förhållandena i nätverkstrafik och därigenom formulera modeller som kan förutsäga flöden i nätverk. Denna studie föreslår en djupinlärningsbaserad metod för att prediktera flöden i nätverk, med hjälp av ett anpassat neuralt nätverk som utnyttjar genvägar i modellens konstruktion (ResNet). Den föreslagna modell-arkitekturen visar sig nöjaktigt kunna förutsäga antalet paket, storlek och varaktighet för flöden med endast den information som är tillgänglig från det första paketet. Dessutom lyckas den föreslagna metoden att överträffa både traditionella maskininlärningsmetoder som linjär regression och beslutsträd, samt konventionella djupa neurala nätverk. Resultaten indikerar att den föreslagna metoden kan förutsäga den allmänna storleken på flödens egenskaper med hög noggrannhet, givet att IP-adresser är tillgängliga.

Neste trabalho é proposto um método de carregamento incremental dos fluxos veiculares para a estimação de matriz O-D sintética. A principal motivação para o desenvolvimento deste trabalho está na complexidade dos métodos de estimação de matriz sintética pelo método iterativo, que tem conduzido a resultados não satisfatórios. O método de carregamento incremental, ora proposto neste trabalho, pode ser definido como o inverso do método de alocação incremental de viagens aos arcos de uma rede viária. A matriz O-D obtida com este método é o resultado da soma sucessiva das sub-matrizes estimadas através das parcelas dos fluxos observados nos arcos da rede. Este método pode ser aplicado em qualquer tipo de rede viária: congestionada ou não congestionada e com ou sem continuidade de fluxo. Para verificação do desempenho do método foram realizados testes experimentais, nos quais os resultados obtidos foram comparados com os valores observados. O desempenho do método incremental não se mostrou satisfatório. Assim, com o intuito de verificar a sua utilidade como um método alternativo os mesmos dados foram processados através de dois métodos iterativos. Os resultados mostraram que os erros são ainda maiores do que os obtidos pelo método proposto. A principal conclusão deste trabalho é que o método incremental pode ser usado como uma alternativa aos métodos iterativos. / A synthetic O-D matrix estimation method, based on incremental loading of traffic flow, was proposed in this work. This research was motivated because of the complexity of the iterative methods for synthetic matrix estimation that might produce bias accumulation in the results. The referred incremental loading method could be defined as the inverse of the incremental assignment method for trips to links of a traffic network. The O-D matrix is gathered by successively summing the sub-matrices obtained from parts of the traffic flow counted on the links of the traffic network. This method could be applied to any traffic networks: congested or uncongested and with or without volumetric continuity. As a part of verification proceeding, several experimental tests were carried out to evaluate the method performance. In these tests the estimated results were compared with the real values. These results show that the incremental loading method performance was not satisfactory. Thus, aiming to test the utility of the incremental method, a second round of experimental tests was conducted to evaluate two iterative methods. In these second round tests, the results show that theses methods performance was worse those of the incremental method. The main conclusion of this work is that the incremental loading method for synthetic matrices estimation could be used as an alternative to the iterative methods.

Neste trabalho é proposto um método de carregamento incremental dos fluxos veiculares para a estimação de matriz O-D sintética. A principal motivação para o desenvolvimento deste trabalho está na complexidade dos métodos de estimação de matriz sintética pelo método iterativo, que tem conduzido a resultados não satisfatórios. O método de carregamento incremental, ora proposto neste trabalho, pode ser definido como o inverso do método de alocação incremental de viagens aos arcos de uma rede viária. A matriz O-D obtida com este método é o resultado da soma sucessiva das sub-matrizes estimadas através das parcelas dos fluxos observados nos arcos da rede. Este método pode ser aplicado em qualquer tipo de rede viária: congestionada ou não congestionada e com ou sem continuidade de fluxo. Para verificação do desempenho do método foram realizados testes experimentais, nos quais os resultados obtidos foram comparados com os valores observados. O desempenho do método incremental não se mostrou satisfatório. Assim, com o intuito de verificar a sua utilidade como um método alternativo os mesmos dados foram processados através de dois métodos iterativos. Os resultados mostraram que os erros são ainda maiores do que os obtidos pelo método proposto. A principal conclusão deste trabalho é que o método incremental pode ser usado como uma alternativa aos métodos iterativos. / A synthetic O-D matrix estimation method, based on incremental loading of traffic flow, was proposed in this work. This research was motivated because of the complexity of the iterative methods for synthetic matrix estimation that might produce bias accumulation in the results. The referred incremental loading method could be defined as the inverse of the incremental assignment method for trips to links of a traffic network. The O-D matrix is gathered by successively summing the sub-matrices obtained from parts of the traffic flow counted on the links of the traffic network. This method could be applied to any traffic networks: congested or uncongested and with or without volumetric continuity. As a part of verification proceeding, several experimental tests were carried out to evaluate the method performance. In these tests the estimated results were compared with the real values. These results show that the incremental loading method performance was not satisfactory. Thus, aiming to test the utility of the incremental method, a second round of experimental tests was conducted to evaluate two iterative methods. In these second round tests, the results show that theses methods performance was worse those of the incremental method. The main conclusion of this work is that the incremental loading method for synthetic matrices estimation could be used as an alternative to the iterative methods.

Apresenta-se um framework que permite a caracterização, conformação e escalonamento de todas as categorias de serviços A TM, dentro do ambiente de usuário (CEP). Propõem-se um modelo de tráfego otimizado para este ambiente que oferece condições de garantias de QoS individuais para fluxos VBR, tempo real ou não, CBR, ABR e UBR. O modelo proposto, denominado de PCSTS (priority class services traffic shaping), inclui um módulo que permite a obtenção do descritor de tráfego de fluxos desconhecidos. Os fluxos são conformados de forma individual, e em tempo real, segundo um conformador baseado num algoritmo de escalonamento virtual duplo baseado no GCRA do ITU/ ATM Forum. Para a multiplexação dinâmica dos diferentes fluxos do ambiente de usuário, no enlace de acesso, é proposto um algoritmo de escalonamento EDD, modificado através de um sinal de realimentação entre conformador e escalonador, e desta forma tem-se condições de oferecer garantias de limite de atraso e jitter para fluxos individuais de serviços rt-VBR ou nrt-VBR. São apresentadas simulações, tanto da caracterização dos fluxos desconhecidos como da arquitetura do escalonador, que comprovam o comportamento esperado do modelo de tráfego PCSTS. / We present a framework that enhances the characterization, shaping and scheduling of ATM traffic flows for ali ATM service categories in the customer prernise equipment (CPE) environment. We propose an optirnized traffic model, capable to guarantee QoS parameters at per connection levei for VBR services, real time or not, CBR, ABR and UBR services. The model proposed, narned PCSTS (priority classes services traffic shaping), includes a module that performs the characterization of services with unknown traffic descriptor. The flows are shaped individually, in real time, through a shaper based on the double discrete time scheduling algorithm of the GCRA from ITU/ ATM Forum. For the dynarnic multiplexing of the different service flows , at the link levei, a modified EDD scheduling algorithm is proposed with a feedback signal between the scheduling and shaping modules. As a result, it is demonstrated that it is possible to grant jitter and delay bounds for the rt-VBR and nrt-VBR services. We present simulations of the traffic characterization module and of the scheduling architecture, for different services, which confirms the expected behavior of the PCSTS model.

Apresenta-se um framework que permite a caracterização, conformação e escalonamento de todas as categorias de serviços A TM, dentro do ambiente de usuário (CEP). Propõem-se um modelo de tráfego otimizado para este ambiente que oferece condições de garantias de QoS individuais para fluxos VBR, tempo real ou não, CBR, ABR e UBR. O modelo proposto, denominado de PCSTS (priority class services traffic shaping), inclui um módulo que permite a obtenção do descritor de tráfego de fluxos desconhecidos. Os fluxos são conformados de forma individual, e em tempo real, segundo um conformador baseado num algoritmo de escalonamento virtual duplo baseado no GCRA do ITU/ ATM Forum. Para a multiplexação dinâmica dos diferentes fluxos do ambiente de usuário, no enlace de acesso, é proposto um algoritmo de escalonamento EDD, modificado através de um sinal de realimentação entre conformador e escalonador, e desta forma tem-se condições de oferecer garantias de limite de atraso e jitter para fluxos individuais de serviços rt-VBR ou nrt-VBR. São apresentadas simulações, tanto da caracterização dos fluxos desconhecidos como da arquitetura do escalonador, que comprovam o comportamento esperado do modelo de tráfego PCSTS. / We present a framework that enhances the characterization, shaping and scheduling of ATM traffic flows for ali ATM service categories in the customer prernise equipment (CPE) environment. We propose an optirnized traffic model, capable to guarantee QoS parameters at per connection levei for VBR services, real time or not, CBR, ABR and UBR services. The model proposed, narned PCSTS (priority classes services traffic shaping), includes a module that performs the characterization of services with unknown traffic descriptor. The flows are shaped individually, in real time, through a shaper based on the double discrete time scheduling algorithm of the GCRA from ITU/ ATM Forum. For the dynarnic multiplexing of the different service flows , at the link levei, a modified EDD scheduling algorithm is proposed with a feedback signal between the scheduling and shaping modules. As a result, it is demonstrated that it is possible to grant jitter and delay bounds for the rt-VBR and nrt-VBR services. We present simulations of the traffic characterization module and of the scheduling architecture, for different services, which confirms the expected behavior of the PCSTS model.