The IEEE Power over Ethernet (PoE) protocol is currently the most used standard for allowing Power Sourcing Equipment (PSE) to distribute power to Powered Devices (PD) through standard Ethernet cables, permitting devices to be added to a network without needing extra power cables.There exist many PSE manager circuits which can be used to easily distribute power to a handful of PDs, but a problem arises when more are desired, since very few of these circuits can communicate with other circuits, and thus can't coordinate power distribution with each other.The goal of this work was to discover the feature set of some popular PSE managers, discover any real-time requirements they may demand of a host, and to then use this knowledge to create an API which could coordinate multiple PSE managers to handle power distribution.The work was carried out with a study of three PSE managers (PD69200, LTC4291, and TPS23880), as well as study and experimentation of the real-time requirements. After this a series of different architectures were created and evaluated using a combination of analysis, questionnaires, and prototype implementation. The work resulted in a software architecture which allows any number of PSE managers of any model to be used simultaneously in a system, with minimal-to-zero considerations needing to be made for different PSE managers. This allows the system hardware to be designed free of any software considerations, and allows the user to operate the system similarly regardless of the underlying architecture.

Today’s robotics technology is often extremely user and location dependable, meaning that only a specific user using a specific hardware or software interface in a specific location can access that technique and alter it. This thesis focuses on an oil-monitoring system called Orilink, with those exact constraints. Orilink is only accessible through a specific computer using a serial port located in a special place. It is neither accessible from elsewhere nor through the cloud. This thesis removes that constraint and enables the access to Orilink independently of the user’s location or software / hardware interface.

The use of peer-to-peer (P2P) applications is growing dramatically, particularly for sharing content such as video, audio, and software. The traffic generated by these applications represents a large proportion of Internet traffic. For the broadband access network providers P2P traffic presents several problems. This thesis identifies the performance and business issues that P2P traffic has on broadband access networks employing the McCircuit separation technique. A mechanism for managing P2P within the access network is proposed. The P2P diversion algorithm aims to manage P2P traffic within the access network based on layer 2 and layer 3 information without employing intrusive layer 7 traffic detection. To solve the contention problem experienced by best effort traffic in the access network, a solution based on the diversion algorithm and on a QoS based traffic classification scheme is proposed. A business model defining the business roles and pricing schemes is presented based on the features offered by the P2P diversion algorithm introducing new opportunities for gaining revenue from P2P traffic for the network service providers and providing better services to users. / Användningen av peer-to-peer (P2P) applikationer ökar dramatiskt, speciellt för spridningen av video, musik, och mjukvara. Trafiken som skapas av dessa program utgör en stor del of trafiken på Internet. För bredbandsaccess operatörer ställer P2P trafik många problem. I detta examensarbete så identifieras både de egenskaper och affärsaspekter som P2P trafiken har på ett bredbandsaccessnät som använder McCircuit som seprationsmekanism för trafiken mellan användare och en mekanism, "peer-to-peer diversion mechansim" (P2PDA), för att hantera P2P trafiken i ett McCircuit baserat accessnät beskrivs. P2PDA algoritmen hanterar P2P trafik i accessnätet baserat på lager 2 och lager 3 information utan att ta hänsyn till applikationslagret (Lager 7). För att få en bra fördelning mellan best-effort trafik och prioriterad trafik så föreslås en lösning baserad på kombinationen av P2PDA och QoS baserad trafik klassificering. Slutligen så defineras en affärsmodell där affärsroller och olika varianter på prissättning för P2P diskuteras baserad på de egenskaper som den förslagna algoritmen medför och den ekonomiska vinst som denna lösning medger.

The domain specific language P4 is a novel initiative which extends the Software-Defined Networking (SDN) paradigm by allowing for data plane programmability. Network virtualisation is a class of network technologies which can be used to abstract the addressing in a network, allowing multiple tenants to utilise the network resources while being agnostic to the underlying network and the other tenants. In other words, twins of tenants using the same addresses can co-exist on the same underlying network. If a twin is a distributed network, it may even be spread out across multiple sites which are connected to a common backbone. In this study, network virtualisation using P4 is evaluated with emphasis on scalability in terms of number of twins and sites. A set of potential network virtualisation technologies are identified and categorised. Based on this categorisation, two variations of network virtualisation are implemented on the P4 capable software switch BMv2 and the performance of both variations are evaluated against the non-P4 solution Linux bridge. Linux bridge was found to yield 451 times more useful bandwidth than the best performing P4 implementation on BMv2, while also learning MAC addresses faster and generating less traffic on the backbone. It is concluded that the performance of network virtualisation implemented and running on BMv2 is worse compared to the non-P4 solution Linux bridge.

The goal of the project was to create independent system, which is able to supervise remote site using Ethernet network. It can measure temperature and check state of door magnetic contact. Gained information are sent over network to the cenral node, which is able to present them on web page. In the project are also discussed important network protocols needed for implementation of the system, hardware used for implementation and software of the system.

Plastic optical fiber (POF) for use in automotive applications is not a new concept and has been used in some vehicles for infotainment media distribution within the Media Oriented Systems Transport protocol. However, the use of POF for the control network’s physical layer is a concept that has not been implemented in automotive applications. Many aspects of a vehicle can be improved by implementing POF as the physical backbone for the control network. Currently, the Controller Area Network (CAN) is used as the primary backbone control network protocol for most automobiles as it is inexpensive and reliable. However, CAN is limited to 500 kbps in most vehicles and is easily accessible. Ethernet may provide the improvements of speed and security needed in today’s feature rich and connected vehicles. The feasibility of implementing Ethernet over POF as the control network for automotive applications is the topic of this research investigation.

Networks within data centers, such as connections between servers and disk arrays, need lossless flow control allowing all packets to move quickly through the network to reach their destination. This paper proposes a new algorithm for congestion control to satisfy the needs of such networks and to answer the question: Is it possible to provide circuit-less reliability and flow control in an Ethernet network? TCP uses an end-to-end congestion control algorithm, which is based on end-to-end round trip time (RTT). Therefore its flow control and error detection/correction approach is dependent on end-to-end RTT. Other approaches utilize specialized data link layer networks such as InfiniBand and Fibre Channel to provide network reliability. The algorithm proposed in this thesis builds on the ubiquitous Ethernet protocol to provide reliability at the data link layer without the overhead and cost of the specialized networks or the delay induced by TCP’s end-to-end approach. This approach requires modifications to the Ethernet switches to implement a back pressure based flow control algorithm. This back pressure algorithm utilizes a modified version of the Random Early Detection (RED) algorithm to detect congestion. Our simulation results show that the algorithm can quickly recover from congestion and that the average latency of the network is close to the average latency when no congestion is present. With correct threshold and alpha values, buffer sizes in the network and on the source nodes can be kept small to allow little needed additional hardware to implement the system.

Background: The train communication network is vulnerable to intrusion assaultsbecause of the openness of the ethernet communication protocol. Therefore, an intru-sion detection system must be incorporated into the train communication network.There are many algorithms available in Machine Learning(ML) to develop the Intru-sion Detection System(IDS). Majorly, depending on the accuracy and execution timeof the algorithm, it is decided as the best. Performance metrics like F1 score, preci-sion, recall, and support are compared to see how well the algorithm fits the modelwhile training. The following thesis will detect the anomalies in the Train ControlManagement System(TCMS) and then the comparison of various algorithms will beheld in order to declare the accurate algorithm. Objectives: In this thesis work, we aim to research anomaly detection in a train’sreal-time ethernet using an IDS. The main objectives of this thesis include per-forming Principal Component Analysis(PCA) and feature selection using RandomForest(RF) for simplifying the complexity of the dataset by reducing dimensionalityand extracting significant features. Followed by, choosing the most consistent algo-rithm for anomaly detection from the selected algorithms by evaluating performanceparameters, especially accuracy and execution time after training the models usingML algorithms. Method: This thesis necessitates one research methodology which is experimen-tation, to answer our research questions. For RQ1, experimentation will help usgain better insights into the dataset to extract valuable and essential features as apart of feature selection using RF and dimensionality reduction using PCA. RQ2also uses experimentation because it provides better accuracy and reliability. Afterpre-processing, the data will be used to train the algorithms and will be evaluatedusing various methods. Results: In this study, we have analysed data using EDA, reduced dimensionalityand feature selection using PCA and RF algorithm respectively. We used five su-pervised machine learning methods namely, Support Vector Machine(SVM), NaiveBayes, Decision Tree, K-nearest Neighbor(KNN), and Random Forest(RF). Aftertesting and utilizing the "KDDCup 1999" pre-processed dataset from the Universityof California Irvine(UCI) ML repository, Decision Tree model has been concludedas the best-performing algorithm with an accuracy of 98.89% in 0.098 seconds, incomparison to other models. Conclusions: Five models have been trained using the five ML techniques foranomaly detection using an IDS. We concluded that the decision tree trained modelhas optimal performance with an accuracy of 98.89% and time of 0.098 seconds

No description available.

Unit testing is a fundamental practice in software development and the goal is to create a test suite that tests the robustness of the software. It is challenging to create a test suite that covers every possible input to a system which can lead to security flaws not being detected. Fuzz testing is a technique that creates randomly generated, or fuzzy, input with the goal to uncover these areas of the input space potentially missed by the unit test suite.  EtherNet/IP is an industrial communications protocol built on top of the TCP/IP suite. HMS Anybus develops hardware to use in secure networks in industrial settings utilizing the EtherNet/IP protocol.  This report outlines the development of a Scapy-based fuzz testing tool capable of testing the implementation of the protocol on HMS devices. Additionally we propose a strategy for how the tool can be deployed in future testing. The resulting fuzz testing tool is capable of creating packets containing selected commands’ encapsulation headers and layering them with command specific data fields. These packets can be filled with static or fuzzy input depending on user configuration. The tool is implemented with the intention of providing HMS the capability for conducting fuzz testing. The report mentions multiple improvements that can be made using A.I. assisted generation of test cases and how the tool can be scaled in the future. This thesis project is a proof of concept that using Scapy to create a fuzz testing tool tailored to the EtherNet/IP protocol is possible.