Made available in DSpace on 2014-06-12T16:00:28Z (GMT). No. of bitstreams: 2 arquivo6669_1.pdf: 1298139 bytes, checksum: 82c0aa9def52f663c245e3f57be952ef (MD5) license.txt: 1748 bytes, checksum: 8a4605be74aa9ea9d79846c1fba20a33 (MD5) Previous issue date: 2007 / Conselho Nacional de Desenvolvimento Científico e Tecnológico / The Transmission Control Protocol (TCP) is responsible for supplying reliable data transport service on the TCP/IP stack and for carrying most than 90% of all Internet traffic. In addition, the stability and efficiency of the actual TCP congestion control mechanisms have been extensively studied and are indeed well known by the networking community. However, new Internet applications and functionalities continuously modify its traffic characteristics, demanding new research in order to adapt TCP to the new reality of the Internet. In particular, a traffic phenomenon known as "mice and elephants" has been motivating important researches around the TCP. The main point is that the standard TCP congestion control mechanisms were designed for elephants leading small flows to experience poor performance. This is caused by the exponential behavior of Slow Start which often causes multiple packet losses due their aggressive increase. This work examines minutely the problems caused by the standard TCP congestion control to mice flows as well as it studies the most important proposals to solve them. Thus, based on such research studies, a modified TCP startup mechanism was proposed. The Burst TCP (B-TCP) is an intuitive TCP modification that employs a responsive congestion window growth scheme based on the current window size, to improve performance for small flows. Moreover, B-TCP is easy to implement and requires TCP adjustment at the sender side only. Simulation experiments show that B-TCP can significantly reduce both transfer times and packet losses for small flows without causing damage to large flows

Orientador: Nelson L. S. da Fonseca / Dissertação (mestrado profissional) - Universidade Estadual de Campinas, Instituto de Computação / Made available in DSpace on 2018-08-04T21:06:39Z (GMT). No. of bitstreams: 1 Godoy_HenriAlvesde_M.pdf: 2153990 bytes, checksum: 515d882e4e703a973dcba331baf6126e (MD5) Previous issue date: 2004 / Resumo: O oferecimento crescente dos serviços de banda larga tem tornado cada vez mais comum o uso de aplicações multimídia na Internet, e o número de transmissões de áudio e vídeo tem crescido nos dias atuais. No entanto, tais fluxos tipicamente utilizam o protocolo de transporte UDP, que se comporta de forma injusta em relação ao TCP. As conexões TCP procuram sempre se adaptar às condições da rede, compartilhando a largura de banda de maneira justa, aproveitando a banda disponível. As aplicações baseadas em UDP, quando competem com tráfegos TCP, tendem a ocupar toda a banda disponível, pois não reagem a nenhuma notificação de congestionamento da rede. O presente trabalho visa apresentar mecanismos de controle de congestionamento e regras para calcular o nível de congestionamento da rede; pretende, ainda, adaptar os fluxos UDP para coexistirem com as conexões TCP, transformando o tráfego UDP das aplicações multimídia em amigáveis ao TCP, tentando evitar, desse modo, um colapso na rede e a inanição do tráfego TCP / Abstract: The increasing number of multimedia applications over the Internet typically uses UDP as transport protocol. UDP does not implement any congestion control mechanism and, thus, does not reduce its sending rate in a congested network, making unfair the share of bandwidth with TCP flow. The present work presents existing mechanisms for adapting UDP senders to share the medium with TCP senders in a fair way. These mechanisms are called TCP friendly and they aim at avoiding network collapse and TCP starvation / Mestrado / Redes de Computadores / Mestre Profissional em Computação

Orientador: Ricardo de Oliveira Anido / Dissertação (mestrado) - Universidade Estadual de Campinas. Instituto de Computação / Made available in DSpace on 2018-07-26T02:06:05Z (GMT). No. of bitstreams: 1 Simonaci_AldecirVieira_M.pdf: 12205330 bytes, checksum: f87ac1df2f464643b1d97f075970acd3 (MD5) Previous issue date: 2000 / Resumo: O crescimento da Internet e a necessidade atual de transporte de diferentes tipos de mídias por suas conexões apresentam novas exigências aos protocolos de roteamento existentes Diversos estudos têm sido realizados para adequar as redes TCP/IP aos novos requisitos de utilização da banda passante, visando a melhorar a confiabilidade, a velocidade e a distribuição da carga. O roteamento multicast surgiu como o resultado do desenvolvimento de soluções mais eficientes para problemas relacionados com a redução de tráfego, propiciando maior flexibilidade e melhor escalabilidade. Este trabalho apresenta um estudo dos protocolos de roteamento multicast utilizados sobre as redes TCP/IP Algumas modificações foram introduzidas no funcionamento dos processos de dilúvio e podas do Protocolo de Roteamento Multicast Vetor de Distância e do Protocol Independent Multicast - modo denso As modificações foram analisadas simulando os protocolos em um simulador de redes. / Abstract: The growth of the Internet and the current need to transport different types of medias through its connections present new demands to the existent routing protocols Several studies have been accomplished to adapt the TCP/TP networks to the new requirements of use of the bandwidth, seeking to improve reliability, speed and load distribution. Multicast routing appeared as a result of the development of more efficient solutions for problems related with the reduction of network traffic, allowing more flexibility and better scalability. This work presents a study of multicast routing protocols used on the TCP/IP networks. Some modifications were introduced in the operation of the flooding and prunings processes of the Distance Vector Multicast Routing Protocol and Protocol Independent Multicast - Dense Mode The modifications were analysed by simulating the protocols on a network simulator. / Mestrado / Mestre em Ciência da Computação

Orientador: Paulo Licio de Geus / Dissertação (mestrado) - Universidade Estadual de Campinas, Instituto de Computação / Made available in DSpace on 2018-07-27T10:46:39Z (GMT). No. of bitstreams: 1 Lima_MarceloBarbosa_M.pdf: 3411926 bytes, checksum: 181bfa6a35bb2e4d536f334d0778de22 (MD5) Previous issue date: 2000 / Resumo: As tecnologias mais tradicionais de firewall - filtros de pacotes e agentes proxy - apresentam algumas limitações que dificultam o suporte a certos serviços e protocolos TCP/IP, que neste trabalho são denominados de "serviços inseguros". Em geral, a única alternativa oferecida, para eliminar os riscos e prover um ambiente mais seguro, é o simples bloqueio incondicional de todo tráfego para estes serviços. Serviços baseados em RPC e UDP são exemplos importantes desta classe. Entretanto, um grande número de novos e interessantes serviços-em sua maioria baseados em UDP e portanto inseguros-têm surgido muito rapidamente com o sucesso da Internet. Faz-se necessário, portanto, o uso de novas estratégias e/ou tecnologias para permitir a utilização de tais serviços sem comprometer a segurança do sistema. Este trabalho tem o objetivo de mostrar que filtros de pacotes com estados são uma excelente solução para a provisão destes serviços inseguros. Como aplicação do estudo realizado neste trabalho, são apresentados detalhes da implementação de um filtro de pacotes com estados capaz de suportar serviços baseados em RPC e que será oficialmente incorporado ao kemel do Linux (série 2.4). / Abstract: Most tradicional firewall technologies, such as packet filters and proxy agents, present some limitations that make it difficult to support some TCP/IP services and protocols, hereafter called "unsecure services". In general, the only alternative offered to eliminate the risks and to provide a more secure environment, is simply to unconditionally deny all traffic for these services. UDP and RPC based services are important examples of this class of services. However, a great deal of new and interesting services-mainly UDP based and as such unsecure-have appeared lately with the Internet's success. It is becoming necessary, therefore, the use of newer strategies and/or technologies to allow these services without jeopardizing the security of the network. The goal od this work is to show that stateful packet filters are an excellent solution for the provision of such unsecure services. As an application of this study, details of a stateful packet filter implementation are presented, one that is able to support RPC-based services, and that will be officially incorporated into the Linux kernel (2.4 series). / Mestrado / Mestre em Ciência da Computação

Orientador: Mauricio F. Magalhães / Dissertação (mestrado) - Universidade Estadual de Campinas, Faculdade de Engenharia Eletrica e de Computação / Made available in DSpace on 2018-07-28T13:21:07Z (GMT). No. of bitstreams: 1 Pagani_CarlosEduardo_M.pdf: 3181301 bytes, checksum: 268feb52cd4cd7037a9ef0256fdef0fe (MD5) Previous issue date: 2000 / Resumo: Novas aplicações estão surgindo para integrar o uso de dados, voz e vídeo na Internet. Essas aplicações requisitam de qualidade de serviço (QoS - Quality of Service) na transmissão dessas informações. Isso motivou o desenvolvimento de serviços adicionais à arquitetura TCP/IP (Transfer Control Protocol / Intemetworking Protocol) para garantir QoS. Em paralelo, a tecnologia ATM (Asyncronous Trasfer Mode) é um importante meio de transmissão em redes de alta velocidade devido a capacidade de transmitir dados, voz e vídeo integrados com garantia de QoS. Esta tese aborda as questões referentes ao uso da tecnologia ATM para suportar os novos serviços IP. Neste trabalho um modelo de comutação e integração IPI A TM é proposto e avaliado por meio de simulação de algumas de suas características básicas / Abstract: New Internet applications integrates data, voice and video forwarding. This applications need Quality of Service (QoS) to control their traffic. Additional services in the TCP/IP (Transfer Control Protocol /Intemetworking Protocol) architecture were proposed to do this. Additionally, A TM is an important transmission mo de in high speed networks due to its capability to send integrated data, voice, and video with QoS. This thesis reports the use of ATM technology to support new IP services. This work proposes a new IP/ATM switching model and simulates the basic characteristics of this architecture / Mestrado / Mestre em Engenharia Elétrica

Made available in DSpace on 2014-06-12T17:42:15Z (GMT). No. of bitstreams: 1 license.txt: 1748 bytes, checksum: 8a4605be74aa9ea9d79846c1fba20a33 (MD5) Previous issue date: 2004 / O fenômeno da ampla disseminação da Internet na última década do século XX trouxe a popularização da tecnologia de redes de computadores e de seu protocolo padrão o TCP/IP. A telefonia convencional é uma delas. Hoje os próprios troncos das companhias telefônicas, são na realidade, backbones de redes TCP/IP sobre ATM ou Frame Relay dedicados exclusivamente ao tráfego de voz, mantendo a qualidade dos serviços em face à uniformidade do tráfego. No tema de convergência para o transporte de diversos serviços de comunicações em uma só infra-estrutura de rede, a Voz sobre IP (VoIP) tem um capítulo à parte. Os conceitos, soluções, protocolos, etc da telefonia IP são aqui representados. A presente dissertação insere-se, por sua vez, no contexto de necessidade de redução de custos operacionais da Prefeitura Municipal de Manaus com telecomunicações, uma vez que já é gasto um valor fixo com a comunicação de dados em sua rede metropolitana

In this thesis we investigate, implement and evaluate a solution for introducing the TransmissionControl Protocol (TCP) into the software of a load generator. The load generator is a simulator usedfor simulating end-user generated activities in the Universal Mobile Telecommunication System(UMTS) network. The purpose of simulating traffic on the network is in this case to verify thefunctionality and robustness of the Radio Network Controller (RNC) node within the UMTSnetwork.TCP is a protocol that provides reliable data transfer over unreliable underlying networkprotocols. It is used as the main transport protocol of the Internet, thus it is also used in the UMTSnetwork in order to provide connectivity for user equipment, such as 3G mobile phones, to servicesover the Internet. For the load generator to be able to produce more realistic traffic scenarios is itdesirable to give it the ability to use TCP.This thesis presents a solution of the problem where an open-source implementation of the TCPfunctionality was chosen, ported to the running platform of the load generator and then tested in asimulated test environment. The choice of the open source implementation of TCP was made byperforming an investigation of available options. In the investigation an open source TCP/IP stackcalled lwIP was chosen. lwIP was then ported to the running platform of the load generator bywrapping and modify the source code. The tests of the ported TCP implementation were made in asimulated test environment with focus on testing basic TCP functionality. The tests showed that theTCP implementation produced provided the basic functionality that was asked for.

The purpose of this thesis is to evaluate and analyze the performance of RED (Random Early Detection) algorithm and our proposed RED algorithm. As an active queue management RED has been considered an emerging issue in the last few years. Quality of service (QoS) is the latest issue in today’s internet world. The name QoS itself signifies that special treatment is given to the special traffic. With the passage of time the network traffic grew in an exponential way. With this, the end user failed to get the service for what they had paid and expected for. In order to overcome this problem, QoS within packet transmission came into discussion in internet world. RED is the active queue management system which randomly drops the packets whenever congestion occurs. It is one of the active queue management systems designed for achieving QoS. In order to deal with the existing problem or increase the performance of the existing algorithm, we tried to modify RED algorithm. Our purposed solution is able to minimize the problem of packet drop in a particular duration of time achieving the desired QoS. An experimental approach is used for the validation of the research hypothesis. Results show that the probability of packet dropping in our proposed RED algorithm during simulation scenarios significantly minimized by early calculating the probability value and then by calling the pushback mechanism according to that calculated probability value. / +46739567385(Rajiv), +46762125426(Mustafa)

This thesis will address a problem concerning availability of information systems at an enterprise within the financial sector and its external suppliers, so called trusted third party. The information system resides on the internal network of the enterprise and must be available to both employees of the enterprise and the trusted third party simultaneously. This contradicts the company policy which disallows third parties access to the internal network. The HAVECA-model introduced in this thesis provides a framework of methods, each solving a sub problem identified in the model. The identified methods are hardening, verification, control and assurance, together supplying a method for continuously securing the internal network against a trusted third party. The HAVECA-model will be applied to a scenario, providing real-world examples of techniques of allowing the trusted third party to a restricted number of servers on the network. / Robert Karlsson e-mail: robert@iron.se Aron Rydquist e-mail: aaron@linux.se

Nowadays a lot of research efforts focus on Mobile Ad-hoc NETworks (MANETs). A MANET is a collection of mobile autonomous nodes, which can move arbitrary, leading to a constantly changing network topology. However, today most of the information is still stored on wired servers. A wired network has a hierarchical topology, while in a MANET the topology is usually flat to allow for nodes to easily change there position in the network. Due to the different topological natures of these systems, interconnectivity is not trivial. To further complicate the situation the Transmission Control Protocol (TCP) is designed for wired networks, in a MANET with different link and route characteristics, as multihop and frequent packet losses, the performance of current TCP proposals drop considerably. The purpose of this report is to give an overview of the current MANET – Internet connectivity situation and to evaluate TCP performance in a hybrid MANET where mobile nodes connect to a wired network through a gateway. The report is divided into two parts. The first theoretical part will evaluate the different routing and mobility problems that occur in a realistic scenario with multiple gateways. Main problems that will be discussed are: path selection, gateway discovery, handover between gateways and address configuration in current solutions. In the last part, a simulation-based evaluation will be made on a simplified scenario where one gateway is linking the two different networks. The simulation will be conducted with ns-2.28, which is extended with support for Uppsala University version of Ad-hoc On-demand Distance Vector (AODVUU) as routing and gateway discovery protocol and TCP AP (Adaptive Pacing) as transport protocol. In the performance evaluation, AODVUU and Destination-Sequenced Distance-Vector Routing (DSDV) combined with TCP Newreno, TCP Vegas and TCP AP will be used. The simulation based evaluation concluded that the best performance was achieved with TCP Vegas in conjunction with AODVUU.