Distributed data storage is a building block for many distributed systems such as mobile file systems, web service replication systems, enterprise file systems, etc. New distributed data storage systems are frequently built as new environment, requirements or workloads emerge. The goal of this dissertation is to develop the science of distributed storage systems by making it easier to build new systems. In order to achieve this goal, it proposes a new policy architecture, PADS, that is based on two key ideas: first, by providing a set of common mechanisms in an underlying layer, new systems can be implemented by defining policies that orchestrate these mechanisms; second, policy can be separated into routing and blocking policy, each addresses different parts of the system design. Routing policy specifies how data flow among nodes in order to meet performance, availability, and resource usage goals, whereas blocking policy specifies when it is safe to access data in order to meet consistency and durability goals. This dissertation presents a PADS prototype that defines a set of distributed storage mechanisms that are sufficiently flexible and general to support a large range of systems, a small policy API that is easy to use and captures the right abstractions for distributed storage, and a declarative language for specifying policy that enables quick, concise implementations of complex systems. We demonstrate that PADS is able to significantly reduce development effort by constructing a dozen significant distributed storage systems spanning a large portion of the design space over the prototype. We find that each system required only a couple of weeks of implementation effort and required a few dozen lines of policy code. / text

M. Tech. Electrical Engineering. / Discusses the question of how to prolong the life of an entire sensor network by reducing the power consumption is the most important issue in wireless sensor networks. Traditional routing schemes have not fulfilled the expectations of an optimised WSN, and as a result, new algorithms are being designed globally. One of these protocols which serve as a basis for this study is the RCRR. In this project, we present a new approach of the Relative Coordinates Rumor Routing algorithm based on a link quality awareness scheme and a modified version of the protocol with better scaling.The benefits of this study are: Proposal of a tuneable algorithm that will not only consume less power compared to the classic RR but also addresses the scalability problem encountered with the original RCRR. Proposal a better trade-off between throughput and average end-to-end delay with the use of the SMAC (compared with Rumor protocol). Avoidance of long delays in queues and saving power on retransmission of packets. Reduction of the wandering of agents during communication. Prolonging the lifetime of the network and thus offering economic, operational, and environmental benefits. Balancing link quality awareness with a topological localisation scheme to achieve better throughput.

M. Tech. Electrical Engineering. / Aims to develope a detection procedure whose objective function is to try as much as possible not to impact heavily on the resource constrained sensor nodes.

M. Tech. Electrical Engineering. / Wireless sensor networks (WSNs) are considered to be the next generation network after the Internet. However the limited energy supply is still a serious restriction for them. Hence energy consumption is a key design criterion for the routing protocol in WSNs. Gradient-based routing (GBR) is an energy efficient routing protocol for WSNs. However, shortcomings exist in the GBR scheme, such as: (1) nodes deliver the message in a point to point manner and do not use the broadcast nature of wireless networks; (2) the nodes which are near the sink will be overused; (3) sinks use flooding to broadcast the interest messages, and hence, many duplicated packets are transmitted. To address these shortcomings, three algorithms are presented according to the problems identified above. Firstly, a competing algorithm GBR-C is proposed. Simulation results showed that GBR-C provided better results than GBR in terms of energy efficiency. Secondly, a refilling algorithm GBR-R is proposed. The simulation results revealed that compared to GBR, network lifetime was prolonged up to 63% with the GBR-R algorithm. Thirdly, a network coding algorithm, GBR-NC, is proposed. Simulation results showed that compared with GBR, GBR-NC could save up to 60% of energy when the network coding scheme N of 4 is used.

We analyze spatio-temporal routing under various constraints specific to multi-robot applications. Spatio-temporal routing requires multiple robots to visit spatial locations at specified time instants, while optimizing certain criteria like the total distance traveled, or the total energy consumed. Such a spatio-temporal concept is intuitively demonstrable through music (e.g. a musician routes multiple fingers to play a series of notes on an instrument at specified time instants). As such, we showcase much of our work on routing through this medium. Particular to robotic applications, we analyze constraints like maximum velocities that the robots cannot exceed, and information-exchange networks that must remain connected. Furthermore, we consider a notion of heterogeneity where robots and spatial locations are associated with multiple skills, and a robot can visit a location only if it has at least one skill in common with the skill set of that location. To extend the scope of our work, we analyze spatio-temporal routing in the context of a distributed framework, and a dynamic environment.

Although the design of Multistage testing (MST) has received increasing attention, previous studies mostly focused on comparison of the psychometric properties of MST with CAT and paper-and-pencil (P&P) test. Few studies have systematically examined the number of items in the routing test, the number of subtests in a stage, or the number of stages in a test design to achieve accurate measurement in MST. Given that none of the studies have identified an ideal MST test design using polytomously-scored items, the current study conducted a simulation to investigate the optimal design for MST using generalized partial credit model (GPCM). Eight different test designs were examined on ability estimation across two routing test lengths (short and long) and two total test lengths (short and long). The item pool and generated item responses were based on items calibrated from a national test consisting of 273 partial credit items. Across all test designs, the maximum information routing method was employed and the maximum likelihood estimation was used for ability estimation. Ten samples of 1,000 simulees were used to assess each test design. The performance of each test design was evaluated in terms of the precision of ability estimates, item exposure rate, item pool utilization, and item overlap. The study found that all test designs produced very similar results. Although there were some variations among the eight test structures in the ability estimates, results indicate that the performance overall of these eight test structures in achieving measurement precision did not substantially deviate from one another with regard to total test length and routing test length. However, results from the present study suggest that routing test length does have a significant effect on the number of non-convergent cases in MST tests. Short routing tests tended to result in more non-convergent cases, and the presence of fewer stage tests yielded more of such cases than structures with more stages. Overall, unlike previous findings, the results of the present study indicate that the MST test structure is less likely to be a factor impacting ability estimation when polytomously-scored items are used, based on GPCM. / text

Mobile ad hoc networks create communication links without the aid of any infrastructure, forwarding packets among mobile nodes. The MANET research community has identified several fundamental challenges, among which the most prominent is discovering an optimal route between two nodes. Existing work has proposed a plethora of routing protocols. Since each protocol implements its own philosophy and algorithm to target a specific purpose, routing protocols in MANETs show very different characteristics. Selecting a particular protocol for an application or deployment environment involves evaluating many complex inter-dependent tradeoffs and can be an overwhelming task for an application designer. However, this decision can have a significant impact on the success of a system in terms of performance, cost, and responsiveness. Emerging distributed applications deployed in MANETs inherently experience highly dynamic situations, which necessitate real-time routing protocol selection in response to varying scenarios. Most of the relevant research in this area relies on simulation studies or empirical analysis to select a routing protocol, requiring an infeasible amount of time and resources for the approaches to be used in real-time decision making. In my dissertation work, I designed the Chameleon framework to facilitate real-time routing protocol decisions based on given application and environmental characteristics. My approach develops analytical models for important network layer performance measures capturing various inter-dependent factors that affect routing protocol behavior. I provide an analytical framework that expresses protocol performance metrics in terms of environment-, protocol-, and application-dependent parameters. This effort has resulted in detailed models for two important metrics: end-to-end delay and throughput. I specify detailed models for the parameters embedded in the models with respect to the ability of network deployers, protocol designers, and application developers to reasonably provide the information. Finally, in a systematic manner, I outline the Chameleon software framework to integrate the analytical models with parameters specified by these three groups of stakeholders. / text

The wireless medium is lossy due to many reasons, such as signal attenuation, multi-path propagation, and collisions. Wireless losses degrade network throughput, reliability, and latency. The goal of this dissertation is to combat wireless losses by developing effective techniques and protocols across different network layers. First, a novel opportunistic routing protocol is developed to overcome wireless losses at the network layer. Opportunistic routing protocols exploit receiver diversity to route traffic in the face of loss. A distinctive feature of the protocol is the performance derived from its optimization can be achieved in real IEEE 802.11 networks. At its heart lies a simple yet realistic model of the network that captures wireless interference, losses, traffic, and MAC-induced dependencies. Then a model-driven optimization algorithm is designed to accurately optimize the end-to-end performance, and techniques are developed to map the resulting optimization solutions to practical routing configurations. Its effectiveness is demonstrated using simulation and testbed experiments. Second, an efficient retransmission scheme (ER) is developed at the link layer for wireless networks. Instead of retransmitting lost packets in their original forms, ER codes packets lost at different destinations and uses a single retransmission to potentially recover multiple packet losses. A simple and practical protocol is developed to realize the idea, and it is evaluated using simulation and testbed experiments to demonstrate its effectiveness. Third, detailed measurement traces are collected to understand wireless losses in dynamic and mobile environments. Existing wireless drivers are modified to enable the logging and analysis of network activity under varying end-host configurations. The results indicate that mobile clients can suffer from consecutive packet losses, or burst errors. The burst errors are then analyzed in more detail to gain further insights into the problem. With these insights, recommendations for future research directions to mitigate loss in mobile environments are presented. / text

Ο ανερχόμενος κλάδος των ασύρματων δικτύων αισθητήρων (Wireless Sensor Networks - WSNs) συνδυάζει δυνατότητα αίσθησης, υπολογισμού και επικοινωνίας σε μια μικρή συσκευή. Ενώ οι δυνατότητες ενός και μόνο κόμβου – αισθητήρα (mote) είναι ελάχιστες, η συνύπαρξη πολλών τέτοιων συσκευών δικτυακά συνδεδεμένων μπορεί να οδηγήσουν σε νέες τεχνολογικές δυνατότητες. Κύριος στόχος αυτών των μικρών συσκευών – κόμβων είναι η δειγματοληψία διάφορων φυσικών μεγεθών, η επεξεργασία αυτών των μεγεθών – μετρήσεων και τέλος, η επικοινωνία του κάθε κόμβου με άλλους με στόχο την μετάδοση αυτών των μετρήσεων – πληροφορίας ασφαλώς και αξιόπιστα. Τα κύρια χαρακτηριστικά αυτών των κόμβων όπως, μικρό μέγεθος, χαμηλό κόστος, αυτονομία και δυνατότητα επικοινωνίας είναι που τα κάνουν τόσο ευέλικτα προσδίδοντας έτσι τη δυνατότητα για ανάπτυξη και υλοποίηση χιλιάδων εφαρμογών. Τα ασύρματα δίκτυα αισθητήρων μπορούν να συμβάλουν σε κρίσιμες εφαρμογές που δεν περιορίζονται αποκλειστικά στον ερευνητικό και επιστημονικό χώρο αλλά επεκτείνονται και στον χώρο της υγείας και της ασφάλειας, όπως παρακολούθηση και επίβλεψη ασθενών, πρόβλεψη και αντιμετώπιση φυσικών καταστροφών, οικιακές εφαρμογές, και στοχεύουν στην βελτίωση της ποιότητας ζωής του ατόμου και της κοινωνίας ως σύνολο. Στο πλαίσιο της παρούσας διπλωματικής εργασίας έγινε προσπάθεια προσέγγισης μιας ανεξερεύνητης έως τώρα σε βάθος εφαρμογής Ασυρμάτων Δικτύων Αισθητήρων, των λεγόμενων «έξυπνων» πόλεων, ως προς τις επικοινωνιακές απαιτήσεις και τους περιορισμούς που θέτει και μια προσπάθεια καταγραφής και αξιολόγησης της απόδοσης και της συμπεριφοράς WSN τεχνολογιών υλικού και επικοινωνιακών τεχνικών σε σχέση με τις απαιτήσεις και τους περιορισμούς αυτούς. Παρουσιάζονται πειραματικά δεδομένα από μετρήσεις που διεξήχθησαν σε εξωτερικό περιβάλλον και εξάγονται χρήσιμα συμπεράσματα σχετικά με μελλοντικές κατευθύνσεις. / The increasingly interesting research field of Wireless Sensor Networks combines the possibility of sensing, computing and networking in a small device. While the capabilities of a single sensor node are limited, the combination of multiple such devices communicating over a network may lead to new technological possibilities and challenges. The main goal of these device nodes is the sampling of several physical values, their processing and the communication of the results to other nodes in a safe and reliable way. The main characteristics of wireless sensor nodes such as the small size, the low cost, autonomy and communication capacity render them flexible and suitable for the development of thousands of applications. Wireless sensor networks are most popular for critical applications that go beyond the research and scientific domain expanding to health, environment and security application classes, like surveillance, patient monitoring, natural disaster prediction and home applications, targeted to improve the quality of life of a single person and of society in general. In the context of this master thesis, the state of the art of a WSN application has been investigated, namely smart cities application field, with respect to the imposed networking requirements and communicational restrictions as well as the analysis and evaluation of the behavioral performance of WSN hardware and communication technologies as far as these requirements and limits are concerned. Finally, experimental results are presented based on measurements conducted outdoors and useful conclusions are extracted forming a future roadmap.

Στόχος της εργασίας είναι να εμβαθύνει πάνω στο σημαντικό θέμα της ποιότητας υπηρεσίας, εξειδικεύοντας το θέμα για τα ATM switches. Μελετά τους αλγορίθμους που εφαρμόζονται ή λειτουργούν σε θεωρητικό επίπεδο για τα ATM switches και τον τρόπο που επηρεάζουν τις διάφορες παραμέτρους του QoS. Για το λόγο αυτό ορισμένοι αλγόριθμοι εφαρμόζονται σε εξομοιωτή δικτύων και με βάση τη συμπεριφορά τους εκτιμάται η επίδραση των διαφόρων αλγορίθμων στην απόδοση και τη συμπεριφορά των δικτύων. / This research examines closely the important issue of quality of service, specializing the issue on ATM switches. It examines the algorithms that are applied or used on theoretical level on ATM switches and the way they affect the QoS parameters. For this reason some algorithms are applied on a network simulator and according the network’s behavior we estimate the effect of different algorithms on networks’ performance and behavior.