Abstract
dc:descriptionInternet of Things is one of the most promising paradigms in the current decade characterized by the use of smart and self-configured objects, like sensors, actuators, wearables etc., that are connected to a network and exchange data by sensing, reacting to events, and interacting with the environment. In addition unmanned mobile devices, e.g. drones, were introduced to users’ daily life the last decade and become a part of the whole of "objects" participating in the IoT as long as they carry sensing equipment and on-board computing elements. These sensing and computational capabilities enhance the network embedded intelligence and allow complex tasks to be realized in a highly distributed fashion, thus, balancing load across the infrastructure and rendering communications much more energy efficient in newly introduced mobile Internet of Things. Their main characteristic of movement to space and time add a new degree of freedom to the needs of monitoring in IoT creating mobile IoT networks. All the IoT objects generate huge amount of data imposing a great demand on processing in order to transform the data into useful information or services. In this new dynamic landscape, it is necessary to have an adequate architecture that can integrate heterogeneous information streams and provide services with an acceptable quality to the users. The realization of an IoT framework needs to take into account many constraints related to the device (power consumption, network processing, battery lifetime etc.), to the stochastic nature of the underlying network (delay, bandwidth utilization, latency ) and to the middleware overlay that is necessary to fuse big volumes of information streams and deliver a service to the user. Therefore it is needed the use of a resource management architecture that can monitor the performance of the units involved in different layers in an IoT system and decides to take actions based on optimal resource use in order to support reliable delivery of information with an acceptable Quality of Service to the users. This thesis proposes the design of a resource management framework which can monitor with no prior knowledge information streams produced by IoT devices, can predict changes with online mechanisms that can disrupt the performance of the IoT framework and can take actions to retain acceptable Quality Of Service while trying to save resources. The online, time optimized and distributed decision making models are based on Optimal Stopping Theory and Change Detection Theory. Starting from the edge we present our research of a dynamic encoder adaptive to changes in video sequences. This encoder proposes a compressing method in order to minimize the error produced to the transmitted multimedia sequence without harming the content. Afterwards, we study the performance of the underlying wireless network and how the changes in network environment can affect and disrupt the mission of unmanned robotic devices and the telemetry received. We study and analyze a time-optimized decision making model adaptive to network quality changes applied in unmanned vehicles as long as the communication between the devices and the fixed control stations can be obstructed, overloaded or can suffer from high packet loss rate due to network variations. Last but not least a stochastic optimization framework of on-line control unit applied on a data distributed platform is proposed because these platforms are necessary to IoT infrastructures to process the enormous volumes of data exchanged between IoT devices. The findings of such decision making models are promising and solidly supportive to a vast spectrum of real-time and latency-sensitive applications with QoS requirements in IoT environments.
Degree
thesis:*- Grantor dc:publisher
- National and Kapodistrian University of Athens
- Year dc:date
- 2019
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Panagidi, Kyriaki
- Παναγίδη, Κυριακή
Subjects
dc:subject × 18- Δυναμική λήψη απόφασης σε πραγματικό χρόνο
- Ανίχνευση αλλαγής
- Θεωρία βέλτιστης παύσης
- Διαδίκτυο των πραγμάτων
- Ομάδα πλαισίων
- Μη επανδρωμένα οχήματα
- Κατανεμημένη πλατφόρμα μηνυμάτων
- Real-time decision making
- Optimal stopping theory
- Unmanned vehicles
- Group-of-pictures
- Distributed streaming platform
- Change detection
- Internet of things (IoT)
- Φυσικές Επιστήμες
- Επιστήμη Ηλεκτρονικών Υπολογιστών και Πληροφορική
- Natural Sciences
- Computer and Information Sciences
Rights
- Language dc:language
- eng
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/46923