No description available.

Very Short Intermi‹ent Distributed Denial of Service (VSI-DDoS) a‹ttack is a new form of DDoS a‹ttacks with potential to bypass many of the security measures used today and still severely damage the quality of service of web applications in cloud systems. Th‘e att‹acks consists of short bursts of legitimate packets which exploits vulnerabilities in the targeted system. With the growing popularity of using containers instead of Virtual Machines in clouds, this project presents an approach for detecting these a‹ttacks in a container based cloud system. Th‘e approach uses signal processing in the form of Discrete Wavelet Transform (DWT) and recurrent neural networks (RNN) called Long Short Term Memory (LSTM) to detect a‹ttacks. Several experiments have been carried out to evaluate the performance of the proposed approach in a controlled testbed environment and it is shown to perform well with competing approaches.

Th‘is paper addresses the possibility of using version control and fi€xed merge pa‹ttern to ensure eventual consistency for collaborative work tools. Where the version control is upheld by a vector clock and used to detect conflƒicts. Which in turn are solved by rolling back to the last common version before using fi€xed merge patt‹ern to create a new common version. In the fixed merge patt‹ern a score system is used to lose as li‹ttle work as possible. ‘This is then tested against a leader node solution for the collaborative work tools. Which upli‰fts a FIFO consistency to sequential consistency by forcing all updates to go through the leader. The test used is a speed test to see if it is a viable solution or not to use when dealing with weak network connection. Since it would require less messages to be sent over the network compared to a leader node solution.

Digitization of objects in three dimensions, also known as 3D scanning, is becoming increasingly prevalent in all types of fields. Ranging from manufacturing, medicine, and even cultural heritage preservation. Many 3D scanning methods rely on cameras to recover depth information and the accuracy of the resulting 3D scan is therefore dependent on their calibration. The calibration process is, for the end-user, relatively cumbersome due to how the popular computer vision libraries have chosen to implement calibration target detection. In this thesis, we have therefore focused on developing and implementing a new type of calibration target to simplify the calibration process for the end-user. The calibration board that was designed is based on colored circular calibration points which form an M-array, where each local neighborhood uniquely encodes the coordinates. This allows the board to be decoded despite being occluded or partially out of frame. This contrasts the calibration board implemented in most software libraries and toolboxes which consists of a standard black and white checkered calibration board that does not allow partial views. Our board was assessed by calibrating single cameras and high FOV cameras and comparing it to regular calibration. A structured light 3D scanning stereo setup was also calibrated which was used to scan and reconstruct calibrated artifacts. The reconstructions could then be compared to the real artifacts. In all experiments, we have been able to provide similar results to the checkerboard, while also being subjectively easier to use due to the support for partial observation. We have also discussed potential methods to further improve our target in terms of accuracy and ease of use.

No description available.

Segmentation of post-operative glioblastoma is important for follow-up treatment. In this thesis, Fully Convolutional Networks (FCN) are utilised together with attention modules for segmentation of post-operative glioblastoma in MR images. Attention-based modules help the FCN to focus on relevant features to improve segmentation results. Channel and spatial attention combines both the spatial context as well as the semantic information in MR images. P-Net is used as a backbone for creating an architecture with existing bottleneck attention modules and was named attention P-Net. The proposed network and competing techniques were evaluated on a Uppsala University database containing T1-weighted MR images of brain from 12 subjects. The proposed framework shows substantial improvement over the existing techniques.

This work presents the design and implementation of a firmware prototype that allows multiple Internet Of Things (IoT) protocols to coexist such that they share a single radio. 6LoWPAN over Time Slotted Channel Hopping (TSCH) and Bluetooth Low Energy (BLE) are the two network protocols that are explored in this work. 6LoWPAN over TSCH forms a mesh network of low power embedded devices so that they are globally accessible through their IPv6 addresses. BLE is suitable for the direct connectivity between devices in short radio ranges. Sharing a single radio between diverse protocols is challenging since protocols access the radio in different modes and with different time patterns. The proposed architecture uses radio abstraction layer for dynamically sharing the radio based on the priority of the requests. Furthermore, concurrent operation of protocols is achieved with the help of priority based multitasking of Micrium operating system. Finally, the time critical operations of the TSCH and the radio abstraction module are handled with nested interrupts. The implementation of the firmware is evaluated with mainly two types of experiments. Overall, it is observed from the experiments that both protocols can coexist concurrently and can maintain their connections. In one set of experiments, periodic data is sent over both protocols every 5 seconds (with small random jitter). Basically these experiments involve small data traffic on both protocols. It is observed that the protocol with lower radio priority degrades in performance. Approximately 10% of the packets need re-transmissions, and thus have more latency for the protocol with lower radio priority. The second set of experiments involve file transfers over 3 hops using 6LoWPAN, while one of the middle nodes (acting as a router for the file transfer) also carries BLE session comprising of 17 BLE activities. It is observed that the mean file transfer time with BLE session is delayed compared to the no BLE case. The delay is less than the duration of the two active TSCH slots. This architecture is suitable for sharing a single radio where the network traffic is sparse. This architecture allows small embedded devices with a single radio to be accessible across many protocols without any modifications to the protocol standards. Although the performance degrades in the simultaneous operations, the protocols still keep their respective connections and complete their operations.

Power grids utilize large central generating stations which entails the use of long transmission lines to deliver power to consumers. This approach poses some challenges such as line loses and carbon dioxide (CO2) emission into the environment from the burning of fossil fuels from such large generators. There is also little or no availability of electricity in the rural and isolated areas where the supply of power from the national grid may be considered uneconomical. Distributed generation proffers solutions to these challenges by generating power close to the point of consumption. A nanogrid is an important aspect of distributed generation in which electricity is generated for a single building. Nanogrids usually employ renewable sources of power such as solar and wind energy to generate electricity. Hence, reducing carbon dioxide emission. Also due to the versatility of nanogrids, people in the rural areas can generate their own electricity. However, the intermittent supply of power due to the variations of wind speed during the entire course of a day poses a major challenge in the use of nanogrids. This thesis focuses on the study of the interaction of the savonius wind turbine in an existing nanogrid with a particular load so as to have a better understanding of how the wind turbine parameters such as the wind speed and Tip Speed Ratio can affect the power generated from the turbine. This is achieved by modelling and measuring the power absorption of the savonius wind turbine operating in the nanogrid. From the results obtained in this project, it is glaring that the wind speed and Tip Speed Ratio of the wind turbine play a vital role in the total power harvested from the turbine. If the rotor blades spin too slowly, the wind will pass through the gap between the blades and no power will be generated. Whereas when the blades spin too fast, they act like a shield against the wind speed, creating turbulence in the air as they spin and so when the incoming blade arrives too fast, it hits the turbulent air created by the blade before it and thus, no power will also be generated in this situation. Therefore, it is of utmost importance to design the wind turbine with an optimal Tip Speed Ratio to obtain maximum power from the turbine and thus, improve the reliability and efficiency of the nanogrid technology.

We examine a system where remote applications frequently request fresh information from sensor nodes through an edge and cloud based infrastructure. We study a scheduling policy that ensures fresh information while reducing energy consumption. In this work we build a testbed for evaluating the performance of the policy. We present a solution to synchronizing devices over a wireless network. We discuss the challenges of using resource constrained devices in an environment where power is limited. We run experiments to evaluate the timing of the testbed and to see how the scheduling policy performs compared to the typical periodic schedule. Our results when testing the schedule are inline with the findings in the theoretical literature. We find that good timing of the testbed requires extensive communication for adjusting the timers.

This thesis was inspired by the desire to make working in cluttered spaces easier. Laboratories are packed full of instruments and tools that scientists use to carry out experiments; this inevitably leads to an increased risk for human error as well as an often uncomfortable experience for the user. Protocols are used to carry out experiments and other processes where missteps will most likely spoil the entire experiment. How could one improve the overall experience and effectiveness in such environments? That's when the idea of using Augmented Reality(AR) came to mind. The challenge was to be able to follow a protocol using AR. The application would require objects to be tracked in space while working and recognize which state of the process the user was in. Using OpenCV for the Computer Vision aspect of the application and writing the software in C++, it was possible to create a successful proof-of-concept. The final result was an AR application that could track all the objects being used for the example protocol and successfully detect, and warn, when the user had made a mistake while creating a series of bacteria cultures. There is no doubt therefore, that with more time and development, a more polished product is possible. The question that is left to answer nonetheless, is whether such an application can pass a UX evaluation to determine its usability-value for users in a professional environment.