No description available.

No description available.

The objective of this project was to design and test a secure IP-based architecture suitable for short duration transactions. This included the development of a prototype test-bed in which various operating scenarios (such as cryptographic options, various IP-based architectures and fault tolerance) were demonstrated. A solution based on SIP secured with TLS was tested on two IP based architectures. Total time, CPU time and heap usage was measured for each architecture and encryption scheme to examine the viability of such a solution. The results showed that the proposed solution stack was able to complete transactions in reasonable time and was able to recover from transaction processor failure. This research has demonstrated a possible architecture and protocol stack suitable for IP-based transaction networks. The benefits of an IP-based transaction network include reduced operating costs for network providers and clients, as shared IP infrastructure is used, instead of maintaining a separate IP and X.25 network.

No description available.

Wireless Sensor Network (WSN) is an emerging multi-hop wireless network technology, and the greenhouse network monitoring system is one of the key applications of WSNs in which various parameters such as temperature, humidity, pressure and power can be monitored. Here, we aim to study the performance of a simulation-based greenhouse monitoring system. To design the greenhouse monitoring system based on WSN, we have used ZigBee-based devices (end devices, routers, coordinators, and actuators. Our proposed greenhouse monitoring network has been designed and simulated using the network simulator OPNET Modeller.The investigation is split into two; first, the aim is to find the optimal Transmit (Tx) power set out at sensor nodes and second, the focus is on studying how increasing the number of sensor nodes in the same greenhouse network will affect the overall network performance. ZigBee-based greenhouses corresponded to 4 network scenarios and are simulated using OPNET Modeller in which 22 different transmit (Tx) power (22 cases) in Scenario 1 is simulated, scenario 2, 3 and 4 estimated to 63, 126, 189 number of sensor nodes respectively. Investigating the performance of the greenhouse monitoring network performance metrics such as network load, throughput, packets sent/received and packets loss are considered to be evaluated under varied transmit (Tx) power and increasing number of sensor nodes. Out of the comprehensive studies concerning simulation results for 22 different transmit (Tx) power cases underlying the greenhouse monitoring network (Scenario1), it is found that packets sent/received and packets loss perform the best with the transmitted (Tx) power falling in a range of 0.9 mWatt to 1.0 mWatt while packet sent/received and packet loss are found to perform moderately with the transmitted (Tx) power values that lie in a range of 0.05 mWatt to 0.8 mWatt. Less than 0.05 mWatt and greater than 0.01 microWatt Tx power experience, the worst performance in terms of particularly packet dropped case. For instance, in the case of the packet dropped (not joined packet, i.e., generated at the application layer but not able to join the network due to lack of Tx power), with a Tx power of 0.01 mWatt, 384 packets dropped with a Tx power of 0.02 and 0.03 mWatt, 366 packets dropped, and with a Tx power of 0.04 and 0.05, 336 packet dropped.While increasing the number of sensor nodes, as in scenario 2, 3 and 4, dealing with sensor nodes 63, 126 and 189 correspondingly, the MAC load, MAC throughput, packet sent/received in scenario 2 are found to perform better than that of scenario 3 and scenario 4, while packet loss in scenarios 2, 3 and 4 appeared to be 15%, 12% and 83% correspondingly.

Fundamental problems in graph theory are of four types existence, construction, enumeration and optimization problems. Optimization problems lie at the interface between computer science and the field of operations research and are of primary importance in decision-making. In this thesis, two optimization problems are studied: the edge-integrity of networks and the network flow problem. An implementation of the corresponding algorithms is also realized.The edge integrity of a communication network provides a way to assess the vulnerability of the network to disruption through the destruction or failure of some of its links. While the computation of the edge-integrity of graphs in general has been proven to be NPcomplete, a recently published paper was devoted to a good algorithm using a technique of edge separation sequence for computing the edge integrity of trees. The main results of this paper will be presented and an implementation of this algorithm is achieved.The network flow problem models a distribution system in which commodities are flowing through an interconnected network. The goal is to find a maximum feasible flow and its value, given the capacity constraints for each edge. The three majors algorithms for this problem (Ford -Fulkerso n, Edmonds-Karp method, MPKM algorithm) are discussed, their complexities compared and an implementation of the Ford-Fulkerson and the MPKM algorithms is presented. / Department of Computer Science

This thesis investigates the possibility to analyse a mobile network with sequences of UERC switches specific to each user equipment. An UERC is essentially a channel that carries information and a user equipment connects to different UERCs depending on whether they want to talk and/or send data with different qualities. As a major player in the mobile technology industry, Ericsson strives to optimise the use of the UERCs and are looking for an automated way to detect changes. The first task was to identify and retrieve the required events from the network log files in order to create the UERC sequences. As a way to test the thesis assumption, and give a proof of concept, two different data sets were analysed were changes had been made to the network settings that should have affected the UERC sequences. With the use of n-grams, Markov chains and Bayesian Estimation testing, the changes could be identified and the thesis assumption could be confirmed - UERC sequences provides a possible way of analysing a mobile network.

Modern and future many-core systems represent large and complex architectures. The communication fabrics in these large systems play an important role in their performance and power consumption. Current simulation methodologies for evaluating networks-on-chip (NoCs) are not keeping pace with the increased complexity of our systems; architects often want to explore many different design knobs quickly. Methodologies that trade-off some accuracy but maintain important workload trends for faster simulation times are highly beneficial at early stages of architectural exploration. We propose a synthetic traffic generation methodology that captures both application behaviour and cache coherence traffic to rapidly evaluate NoCs. This allows designers to quickly indulge in detailed performance simulations without the cost of long-running full system simulation but still capture a full range of application and coherence behaviour. Our methodology has an average (geometric) error of 10.9% relative to full system simulation, and provides 50x speedup on average over full system simulation.

One of the impacts of the rising demand for community health services, primarily used by seniors, is that hospitals are often faced with the challenge of having patients finish the acute phase of their treatment and yet are unable to discharge them due to the lack of a bed in a more appropriate community care setting. The frequency of this challenge has led to the designation of “alternative level of care” (ALC) being ascribed to patients who remain in the hospitals due to insufficient capacity downstream. The thesis focuses on a model that seeks to address patient flow through the community care network (CCN) and finding capacity allocation policies for the different facilities that resolves the ALC challenge using scenario analysis. A queuing network model with general routings and nodes’ blocking has been developed and a heuristic approximation method has been employed for solving the model. Blocking probabilities and the number of blocked patients are derived as performance metrics of the CCN. We test the accuracy of the queuing model through a simulation model and the behaviours of the system in different scenarios are investigated in the simulation model and our policy insights and conclusions are provided.

Linear programming (LP) models that contain a (substantial) network structure frequently arise in many real life applications. In this thesis, we investigate two main questions; i) how an embedded network structure can be detected, ii) how the network structure can be exploited to create improved sparse simplex solution algorithms. In order to extract an embedded pure network structure from a general LP problem we develop two new heuristics. The first heuristic is an alternative multi-stage generalised upper bounds (GUB) based approach which finds as many GUB subsets as possible. In order to identify a GUB subset two different approaches are introduced; the first is based on the notion of Markowitz merit count and the second exploits an independent set in the corresponding graph. The second heuristic is based on the generalised signed graph of the coefficient matrix. This heuristic determines whether the given LP problem is an entirely pure network; this is in contrast to all previously known heuristics. Using generalised signed graphs, we prove that the problem of detecting the maximum size embedded network structure within an LP problem is NP-hard. The two detection algorithms perform very well computationally and make positive contributions to the known body of results for the embedded network detection. For computational solution a decomposition based approach is presented which solves a network problem with side constraints. In this approach, the original coefficient matrix is partitioned into the network and the non-network parts. For the partitioned problem, we investigate two alternative decomposition techniques namely, Lagrangean relaxation and Benders decomposition. Active variables identified by these procedures are then used to create an advanced basis for the original problem. The computational results of applying these techniques to a selection of Netlib models are encouraging. The development and computational investigation of this solution algorithm constitute further contribution made by the research reported in this thesis.