This thesis presents research on transport layer behavior in wireless networks. As the Internet is expanding its reach to include mobile devices, it has become apparent that some of the original design assumptions for the dominant transport protocol, TCP, are approaching their limits. A key feature of TCP is the congestion control algorithm, constructed with the assumption that packet loss is normally very low, and that packet loss therefore is a sign of network congestion. This holds true for wired networks, but for mobile wireless networks non-congestion related packet loss may appear. The varying signal power inherent with mobility and handover between base-stations are two example causes of such packet loss. This thesis provides an overview of the challenges for TCP in wireless networks together with a compilation of a number of suggested TCP optimizations for these environments. A TCP modification called TCP-L is proposed. It allows an application to increase its performance, in environments where residual bit errors normally give a degraded throughput, by making a reliability tradeoff. The performance of TCP-L is experimentally evaluated with an implementation in the Linux kernel. The transport layer performance in a 4G scenario is also experimentally investigated, focusing on the impact of the link layer design and its parameterization. Further, for emulation-based protocol evaluations, controlled packet loss and bit error generation is shown to be an important aspect.

Cette thèse s'inscrit dans le domaine de l'expérimentation sur les systèmes distribués, et en particulier de leur test ou de leur validation. À côté des méthodes d'évaluation classiques (modélisation, simulation, plates-formes d'expérimentation comme PlanetLab ou Grid'5000) les méthodes basées sur l'émulation et la virtualisation proposent une alternative prometteuse. Elles permettent d'exécuter l'application réelle à étudier, en lui présentant un environnement synthétique, correspondant aux conditions d'expérience souhaitées : il est ainsi possible, à moindre coût, de réaliser des expériences dans des conditions expérimentales différentes, éventuellement impossibles à reproduire dans un environnement réel. Mais l'utilisation de tels outils d'émulation ne peut se faire sans répondre à des questions sur leur réalisme et leur passage à l'échelle. Dans ce travail, nous utilisons une démarche incrémentale pour construire une plate-forme d'émulation destinée à l'étude des systèmes pair-à-pair à grande échelle. Nous commençons par comparer les différentes solutions d'émulation logicielle de liens réseaux, puis illustrons leur utilisation, notamment en étudiant une application réseau complexe : TUNS, un tunnel IP sur DNS. Nous construisons ensuite notre plate-forme d'émulation, P2PLab, en utilisant l'un des émulateurs réseaux précédemment étudiés, ainsi qu'un modèle de topologies réseaux adapté à l'étude des systèmes pair-à-pair. Nous y proposons une solution légère de virtualisation, permettant un bon rapport de repliement (grand nombre de noeuds émulés sur chaque machine physique). Après avoir validé cette plate-forme, nous l'utilisons pour étudier le protocole de diffusion de fichiers pair-à-pair BitTorrent à l'aide d'expériences mettant en jeu près de 15000 noeuds participants.

<p>This thesis presents research on transport layer behavior in wireless networks. As the Internet is expanding its reach to include mobile devices, it has become apparent that some of the original design assumptions for the dominant transport protocol, TCP, are approaching their limits. A key feature of TCP is the congestion control algorithm, constructed with the assumption that packet loss is normally very low, and that packet loss therefore is a sign of network congestion. This holds true for wired networks, but for mobile wireless networks non-congestion related packet loss may appear. The varying signal power inherent with mobility and handover between base-stations are two example causes of such packet loss. This thesis provides an overview of the challenges for TCP in wireless networks together with a compilation of a number of suggested TCP optimizations for these environments. A TCP modification called TCP-L is proposed. It allows an application to increase its performance, in environments where residual bit errors normally give a degraded throughput, by making a reliability tradeoff. The performance of TCP-L is experimentally evaluated with an implementation in the Linux kernel. The transport layer performance in a 4G scenario is also experimentally investigated, focusing on the impact of the link layer design and its parameterization. Further, for emulation-based protocol evaluations, controlled packet loss and bit error generation is shown to be an important aspect.</p>

<p>Due to the cheap and flexible framework that the underlying IP-technology of the internet provides, IP-networks are becoming popular in more and more contexts. For instance, telecommunication operators have started to replace the fixed legacy telephony networks with IP-networks. To support a smooth transition towards IP-networks, the Stream Control Transmission Protocol (SCTP) was standardized. SCTP is used to carry telephony signaling traffic, and solves a number of problems that would have followed from using the Transmission Control Protocol (TCP) in this context. However, the design of SCTP is still heavily influenced by TCP. In fact, many protocol mechansisms in SCTP are directly inherited from TCP. Unfortunately, many of these mechanisms are not adapted to the kind of traffic that SCTP is intended to transport: time critical message-based traffic, e.g. telephony signaling.In this thesis we examine, and adapt some of SCTP's mechanisms to more efficiently transport time critical message-based traffic. More specifically, we adapt SCTP's loss recovery and message bundling for timely message transfers. First, we propose and experimentally evaluate two loss recovery mechanisms: a packet-based Early Retransmit algorithm, and a modified retransmission timeout management algorithm. We show that these enhancements can reduce loss recovery times with at least 30-50%, in some scenarios. In addition, we adapt the message bundling of SCTP to better support timely message delivery. The proposed bundling algorithm can in some situations reduce the transfer time of a message with up to 70%.In addition to these proposals we have also indentified and reported mistakes in some of the most popular SCTP implementations. Furthermore, we have continously developed the network emulation software KauNet to support our experimental evaluations.</p>

This thesis presents a general simulation framework used for evaluating the performance of bilateral teleoperation systems under consistent and controllable network conditions. A teleoperation system is where an operator uses a master device to control a slave robot through a communication link. The communication link between the master and slave has an important impact on the system performance. Network emulation using ns-2 has been proposed as a way of simulating the communication link. It allows for the network conditions to be controlled and for repeatable results. The proposed setup was used to test the performance of a hydraulic actuator under various conditions of wired and wireless networks. Three control schemes were evaluated using various combinations of time delay and packet loss. The system was also tested simulating wireless communication between the master and slave to determine the effects of transmission power and distance on the performance of the system.

This thesis presents a general simulation framework used for evaluating the performance of bilateral teleoperation systems under consistent and controllable network conditions. A teleoperation system is where an operator uses a master device to control a slave robot through a communication link. The communication link between the master and slave has an important impact on the system performance. Network emulation using ns-2 has been proposed as a way of simulating the communication link. It allows for the network conditions to be controlled and for repeatable results. The proposed setup was used to test the performance of a hydraulic actuator under various conditions of wired and wireless networks. Three control schemes were evaluated using various combinations of time delay and packet loss. The system was also tested simulating wireless communication between the master and slave to determine the effects of transmission power and distance on the performance of the system.

Due to the cheap and flexible framework that the underlying IP-technology of the internet provides, IP-networks are becoming popular in more and more contexts. For instance, telecommunication operators have started to replace the fixed legacy telephony networks with IP-networks. To support a smooth transition towards IP-networks, the Stream Control Transmission Protocol (SCTP) was standardized. SCTP is used to carry telephony signaling traffic, and solves a number of problems that would have followed from using the Transmission Control Protocol (TCP) in this context. However, the design of SCTP is still heavily influenced by TCP. In fact, many protocol mechansisms in SCTP are directly inherited from TCP. Unfortunately, many of these mechanisms are not adapted to the kind of traffic that SCTP is intended to transport: time critical message-based traffic, e.g. telephony signaling.In this thesis we examine, and adapt some of SCTP's mechanisms to more efficiently transport time critical message-based traffic. More specifically, we adapt SCTP's loss recovery and message bundling for timely message transfers. First, we propose and experimentally evaluate two loss recovery mechanisms: a packet-based Early Retransmit algorithm, and a modified retransmission timeout management algorithm. We show that these enhancements can reduce loss recovery times with at least 30-50%, in some scenarios. In addition, we adapt the message bundling of SCTP to better support timely message delivery. The proposed bundling algorithm can in some situations reduce the transfer time of a message with up to 70%.In addition to these proposals we have also indentified and reported mistakes in some of the most popular SCTP implementations. Furthermore, we have continously developed the network emulation software KauNet to support our experimental evaluations.

In the past decade or so, significant advances were made in the field of Future Internet Architecture (FIA) design. Undoubtedly, the size of Future Internet will increase tremendously, and so will the complexity of its users’ behaviors. This advancement means most of future Internet applications and services can only achieve and demonstrate full potential on a large-scale basis. The development of network testbeds that can validate key design decisions and expose operational issues at scale is essential to FIA research. In conjunction with the development and advancement of FIA, cyber-infrastructure testbeds have also achieved remarkable progress. For meaningful network studies, it is indispensable to utilize cyber-infrastructure testbeds appropriately in order to obtain accurate experiment results. That said, existing current network experimentation is intrinsically deficient. The existing testbeds do not offer scalability, flexibility, and realism at the same time. This dissertation aims to construct a hybrid system of conducting at-scale network studies and experiments by exploiting the distributed computing ability of current testbeds. First, this work presents a synchronization of parallel discrete event simulation that offers the simulation with transparent scalability and performance on various high-end computing platforms. The parallel simulator that we implement is configured so that it can self-adapt for the performance while running on supercomputers with disparate architectures. The simulator could be used to handle models of different sizes, varying modeling details, and different complexity levels. Second, this works addresses the issue of researching network design and implementation realistically at scale, through the use of distributed cyber-infrastructure testbeds. An existing symbiotic approach is applied to integrate emulation with simulation so that they can overcome the limitations of physical setup. The symbiotic method is used to improve the capabilities of a specific emulator, Mininet. In this case, Mininet can be used to run applications directly on the virtual machines and software switches, with network connectivity represented by detailed simulation at scale. We also propose a method for using the symbiotic approach to coordinate separate Mininet instances, each representing a different set of the overlapping network flows. This approach provides a significant improvement to the scalability of the network experiments.

This diploma thesis deals with a development of a network emulator operating in TCP/IP computers networks. This work contains a specification of transmission parameters and an analysis of concurrent systems both proprieties and open-sources. Afterward describes a relalization of the emulator itself. The core of the emulator uses well-known Linux tools tc, netem and iptables. An API of the program is realized by web service and for user interaction has been developed a web user interface. At the end of this work are listed and then discussed measured results of the emulator capabilities.

The increasing scale and complexity of network applications and protocols motivates the need for tools to aid in the understanding of network dynamics at similarly large scales. While current network simulation tools achieve large scale modeling, they do so by ignoring much of the intra-program state that plays an important role in the overall system's behavior. This work presents The Distributed Open Network Emulator, a scalable distributed network model that incorporates application program state to achieve high fidelity modeling. The Distributed Open Network Emulator, or DONE for short, is a parallel and distributed network simulation-emulation hybrid that achieves both scalability and the capability to run existing application code with minimal modification. These goals are accomplished through the use of a protocol stack extracted from the Linux kernel, a new programming model based on C, and a scaled real-time method for distributed synchronization. One of the primary challenges in the development of DONE was in reconciling the opposing requirements of emulation and simulation. Emulated code directly executes in real-time which progresses autonomously. In contrast, simulation models are forced ahead by the execution of events, an explicitly controlled mechanism. Relativistic time is used to integrate these two paradigms into a single model while providing efficient distributed synchronization. To demonstrate that the model provides the desired traits, a series of experiments are described. They show that DONE can provide super-linear speedup on small clusters, nearly linear speedup on moderate sized clusters, and accurate results when tuned appropriately. / Master of Science