No description available.

The mammalian visual system develops to perform many complex tasks that allow us to perceive the natural world. These tasks rely on a dense network of synaptic connections transporting visual information both to and within visual cortex (V1). The laminar organization and functional properties of visual cortical neurons are largely conserved across mammals, and the mouse has been adopted as a model organism to study the development of this cortical circuit. Neurons in each cortical layer must find the correct synaptic partners for the optimal receipt, transfer, and processing of information. The molecular cues guiding the development of these connections, however, are largely unknown. In this thesis, I identify and then examine the role of molecular factors important for synapse formation in layer 2/3 (L2/3) of visual cortex. L2/3 neurons are highly interconnected and fire selectively to a refined set of visual stimuli. The developmental refinement of these visual preferences has been shown to occur in the week following eye opening, corresponding with a period of intense synapse formation and dynamic gene expression in mouse V1. In Chapters II–IV, I use the TU-tagging technique to identify molecular factors enriched L2/3 neurons before and after eye opening and identify several candidate genes with potential functions in synapse formation. In Chapter V, I examine the function of cell adhesion molecules nectin-1 and nectin-3, identified here as enriched in L2/3 visual cortex at eye opening, and previously shown to interact across synaptic junctions. I focus mainly on the effect of nectin-3 (having post-synaptic localization in hippocampus) on post-synaptic dendritic spine densities in developing L2/3 cortical neurons. I show that nectin-3 knockdown further increases spine densities after eye opening, while overexpressing a full length or truncated nectin-3 protein reduces spine densities. I conclude that nectin-3 may have a role in synapse formation following eye opening, and propose a mechanism describing the effects observed. Here, I describe a unique approach for understanding how cell-type specific connections are formed in visual cortex, beginning with the spatiotemporal examination gene expression and followed by the spatiotemporal manipulation of a single gene. This dissertation includes previously published co-authored material.

Network management requires a large amount of knowledge about the network. In particular, knowledge about used network addresses, access time, and topology is useful. In a network composed of managed devices, much of the data necessary can come from simple network management protocol (SNMP) queries. Other data can come from other databases, or analysis of existing data. In particular, layer-two network topology can be determined by analyzing the mac address forwarding tables of layer-two devices. The layer-two topology can be merged with a layer-three topology to generate a complete topology of the network. This information is useless unless it is easily accessible to the network manager; therefore, a simple interface should be used to give access to all of this data.

Thesis (MTech (Information Technology))--Cape Peninsula University of Technology, 2019 / The availability of high-speed/high-volume Data Link Layer (Layer 2) transmission networks fuelled by the implementation of mission critical and performance-intensive technologies, such as Cloud and Data Centre services transmitting sensitive data over the wide area network (WAN) has shifted the attention of hackers, eavesdroppers, cyber-criminals and other malicious attackers to the exploitation of these data transmission technologies. It is argued that security on the current classical technologies that store, transmit and manipulate information on the OSI Layer 2 have historically not been adequately addressed when it comes to secure communication and exchange of information. Quantum teleportation (QT) stemming from quantum communication a branch of quantum information science (QIS) has emerged as a technology that promise unconditional security and providing new ways to design and develop frameworks that operate based on the laws of quantum physics. It is argued that it has a potential to address the data transmission security GAP for OSI layer 2 technologies. This research study aims to propose a framework for the implementation of secure quantum teleportation infrastructures in South Africa. There is currently a lack of generic models and methods to guide the implementation of QT infrastructures that will enable secure transmission of information. A design science research (DSR) was undertaken in order to develop a secure quantum teleportation artefact called (SecureQT-Framework). SecureQT-Framework is a generic model and method that guides the selection and implementation of QT infrastructures motivated by multi-disciplinary domains such as QIS, Quantum Physics, Computer Science as well as information and communication technology (ICT). The DSR process employed a primary DSR cycle with four DSR sub-cycles which involved the awareness and suggestion phase guided by a systematic literature review (SLR), development and evaluation phase guided by Software Defined Network’s OpenFlow, Mininet, Mininet-Wifi and computer simulations for QT using SQUANCH framework. We investigated, examined and collected credible QT techniques and its variant protocols to develop and simulate secure transmission of information over the WAN, We studied their features and challenges. We concluded the study by describing the QT techniques, protocols and implementations that has potential to bridge the security GAP for OSI Layer 2 technologies over the WAN. The results gained were used in the construction of a framework for the implementation of a secure quantum teleportation infrastructure in South Africa. The framework describes the main factors that need to be taken into consideration when implementing quantum teleportation infrastructures.

This report aims to compare different layer 2 tunneling protocols in order to determine what advantages and disadvantages each protocol comes with. Since wireless communication is becoming more used demand for these protocols has risen. The main goal is to help industries, like Hardware Meets Software (HMS), choose the most suitable layer 2 tunneling protocol and give a detailed view of each protocol. To compare the protocols, information about various protocols was gathered as well as other studies done on layer 2 tunneling protocols. A decision was then made on the protocols to focus on, Layer 2 Tunneling Protocol (L2TP), Internet Protocol Security (IPsec), Virtual eXtensible Local-Area Network protocol (VXLAN), Generic Routing Encapsulation protocol (GRE), and Network Virtualization using Generic Routing Encapsulation protocol (NVGRE). Furthermore, tests were done on non-tunneling, L2TP together with IPsec and VXLAN, using two laptops, two gateways as well as a shark trap together with Wireshark to record the traffic. These tests were recorded and documented. The results of the tests were used to calculate the protocol’s latency, throughput and jitter. The background research together with the tests concluded in specific protocol characteristics, with different use cases for each protocol. The tests show that non-tunneling was the fastest communication. The L2TP together with the IPsec had better latency and jitter while VXLAN had better throughput. Using other studies together with our tests, different advantages and disadvantages with each protocol was discussed and it was concluded that deeper research would be needed toconclude which protocol best suits HMS, but that L2TP with IPsec outperformed VXLAN on 2 of the 3 main factors. Finally, the methodology is discussed with the factors of reliability, accuracy, and limitation.

Handoff is a call handling mechanism that is invoked when a mobile node moves from one cell to another. Such movement may lead to degradation in performance for wireless networks as a result of packet losses. A promising technique proposed in this thesis is to apply multicasting techniques aided by mobility prediction in order to improve handoff performance. In this thesis, we present a method that uses a Grey model for mobility prediction and a fuzzy logic controller that has been fine-tuned using evolutionary algorithms in order to improve prediction accuracy. We also compare the self-tuning algorithm with two evolutionary algorithms in terms of accuracy and their convergence times. Our proposed method takes into account signal strengths from the base stations and predicts the signal strength of the next candidate base station in order to provide improved handover performance. The primary decision for mobility prediction is the accurate prediction of signal strengths obtained from the base stations and remove any unwanted errors in the prediction using suitable optimisation techniques. Furthermore, the model includes the procedures of fine-tuning the predicted data using fuzzy parameters. We also propose suitable multicasting algorithms to minimise the reservation of overall network resource requirements during handoff with the mobility prediction information. To be able to efficiently solve the problem, the situation is modelled using a multicast tree that is defined to maintain connectivity with the mobile node, whilst ensuring bandwidth guarantees and a minimum hop-count. In this approach, we have tried to solve the problem by balancing two objectives through putting a weight on each of two costs. We provide a detailed description of an algorithm to implement join and prune mechanisms, which will help to build an optimal multicast tree with QoS requirements during handoff as well as incorporating dynamic changes in the positions of mobile nodes. An analysis of how mobility prediction helps in the selection of potential Access Routers (AR) with QoS requirements - which affects the multicast group size and bandwidth cost of the multicast tree -- is presented. The proposed technique tries to minimise the number of multicast tree join and prune operations. Our results show that the expected size of the multicast group increases linearly with an increase in the number of selected destination AR's for multicast during handoff. We observe that the expected number of joins and prunes from the multicast tree increases with group size. A special simulation model was developed to demonstrate both homogeneous and heterogeneous handoff which is an emerging requirement for fourth generation mobile networks. The model incorporates our mobility prediction model for heterogeneous handoff between the Wireless LAN and a cellular network. The results presented in this thesis for mobility prediction, multicasting techniques and heterogeneous handoff include proposed algorithms and models which aid in the understanding, analysing and reducing of overheads during handoff.

Cette thèse propose une nouvelle méthode pour fournir une connexion réseau à des véhicules au cours de trajets prédéterminés (trains, métros, autobus urbains, etc.). La communication entre le véhicule et l'infrastructure réseau est basée uniquement sur la technologie WiFi. Les contributions de ce travail sont d'une part la conception d'une méthode pour réaliser le handover horizontal (entre bornes WiFi), et d'autre part la modélisation et l'analyse de topologies pour le réseau d'infrastructure (réseau backbone plus réseau d'accès WiFi) déployé sur la trajectoire du véhicule. Dans une première partie, nous proposons une méthode, appelée Spiderman Handover, pour réaliser le handover horizontal d'un réseau en mouvement (embarqué dans le véhicule) et une procédure de mise à jour des informations de routage (couche 2 OSI) lors du handover. Nous évaluons notre proposition par simulation et validons nos résultats par des mesures expérimentales. Dans une deuxième partie, nous étudions théoriquement les paramètres de plusieurs familles de topologies du type Chordal pour le réseau backbone construit sur un réseau d'accès linéaire. A partir de la comparaison de ces paramètres, nous proposons une topologie backbone issue de la combinaison de deux topologies Chordal. Cette topologie fournit un bon compromis entre coût du déploiement, nombre de sauts nécessaires pour atteindre la passerelle du réseau et résilience raisonnable. Enfin, nous évaluons l'intégration de la topologie proposée pour le réseau d'infrastructure avec le système handover par des simulations. Les résultats présentés suggèrent que l'algorithme de handover proposé fonctionne correctement sur le réseau d'infrastructure proposé. Cela permet la garantie d'une connexion continue aux passagers à bord des trains, métros ou autobus urbains.

As the networks become faster and faster, the emerging requirement is to improve the performance of the Intrusion Detection and Prevention Systems (IDPS) to keep up with the increased network throughput. In high speed networks, it is very difficult for the IDPS to process all the packets. Since the throughput of IDPS is not improved as fast as the throughput of the switches and routers, it is necessary to develop new detection techniques other than traditional techniques. In this thesis we propose a rule-based IDPS technique to detect Layer 2-4 attacks by just examining the flow data without inspecting packet payload. Our approach is designed to work as an additional component to existing IDPS as we acknowledge that the attacks at Layer 5 and above require payload inspection. The rule set is constructed and tested on a real network to evaluate the performance of the system.

This work shows and implements possibilities of OSI Layer-2 Network Topology Discovery based on information from Cisco Discovery Protocol, Spanning Tree Protocol and Switches' Filtering Databases. / Diese Arbeit zeigt und implementiert Möglichkeiten zur OSI Layer-2 Netzwerk Topologie Erkennung basierend auf Informationen des Cisco Discovery Protocol, Spanning Tree Protocol und Switch Filter Datenbanken.

This work shows and implements possibilities of OSI Layer-2 Network Topology Discovery based on information from Cisco Discovery Protocol, Spanning Tree Protocol and Switches' Filtering Databases. / Diese Arbeit zeigt und implementiert Möglichkeiten zur OSI Layer-2 Netzwerk Topologie Erkennung basierend auf Informationen des Cisco Discovery Protocol, Spanning Tree Protocol und Switch Filter Datenbanken.