Modern applications produce large quantities of data in the form of logs and events in order to facilitate quick failure diagnosis and mitigation. Special Big Data database systems like Elasticsearch are needed to store and manage these logs. Through such a system, the data is used to query different events and trace down issues that appear in the application. To further improve the efficiency of failure discovery and prevention, the query mechanisms can be extended with Machine Learning techniques to observe anomalies in the infrastructure and automatically alert the managers or administrators. This research project is tracking the development of a scalable Elasticsearch monitoring system. In later stages, the focus will fall on enhancing the system's detection functionality by applying Machine Learning jobs on the stored data.

Given the rate at which mobile devices are adopted by the general public, we are at a point where many of us own a smart phone. Currently, this makes them the most affordable interfaces for augmented reality. However, with respect to augmented reality, smartphones present some limitations: location tracking is, on a large scale, supported only for outdoor environments and the content types that can be displayed to the user are generally limited. We therefore propose a marker-based tracking system which detects QR-codes on the camera capture and overlays rich media obtained from a server. We further analyze the performance implications of such a system, by breaking down the processing into pipelined stages and providing a mechanism of skipping frames from the capture in order to give a fast on-screen response.

The past years have shown an increase in the both number and sophistication of cyber-attacks targeting Windows and Linux operating systems. Traditional network security solutions such as firewalls are incapable of detecting and stopping these attacks. In this paper, we describe our distributed firewall solution Distfw and its integration with a sandbox for malware analysis and detection. We demonstrate the effectiveness and shortcomings of such a solution. We use Cuckoo to perform automated analysis of malware samples and compare the results with the ones from manual analysis. We discover that Cuckoo provides similar results in a considerable amount of time.

Home automation systems are collections of interconnected devices for controlling various functions within a house, such as light control, heating, air conditioning, etc. Mobile devices are ideal in providing a user interface in a home automation system, due to their portability and their wide range of capabilities. They can communicate with a home automation network through an Internet gateway, but cannot directly communicate with devices in the network, as these devices usually implement low power communication protocols, such as ZigBee. We investigate several methods to equip an Android device with a dongle capable of ZigBee communication. We propose a scalable architecture, with three abstraction layers, that scales over multiple communication channels, such as the TCP channel, for communication with the gateway, and the USB channel, for direct communication with devices through the dongle. We test the application with two client devices running Android 4.0 and a mock home automation network consisting of a couple of ZigBee devices and a PC as gateway. We estimate the energy consumption of WiFi transfers over several typical use cases and we conclude that using ZigBee can prove beneficial in some cases, both in terms of functionality and performance.

Cloud Computing is a cutting edge technology for managing and delivering services over the Internet. Map-Reduce is the programming model used in cloud computing for processing large data sets in parallel over huge clusters. In order to increase efficiency, a good task scheduling is needed. Genetic algorithms are very useful and accurate in finding solutions to large scale optimization problems, such as task scheduling. They have gained immense popularity over last few years as a robust and easily adaptable search technique. Hadoop, the open source implementation of Map-Reduce, has several task schedulers available (FIFO, Fair, Capacity Schedulers), but neither one of them is focused on minimizing the global execution time. The goal of this project is to improve Hadoop's functionality by implementing a scheduler based on a genetic algorithm, solving the stated problem.

The advances in Internet innovation have made possible techniques for the conveyance of healthcare. Networking infrastructure and common access can encourage sharing of patient data and clinical information; make the Internet a perfect tool for remote patient observing applications. In this paper, we portray an experimental model designed for monitoring and checking the health condition of the patients based on sensors. The framework depends on e-health sensor shield associated with a cloud platform which gathers the data from the sensors. The sensors measure various parameters, such as a glucometer, airflow and patient position which are transmitted via microcontroller by a gateway to a cloud storage platform. The data collected in the cloud platform is accessible for further handling, for the investigation of some correlations among measured parameters and health state of the patients.