k-core is a notion capturing the cohesiveness of a subgraph in a social network graph. Most of the current research work only consider pure network graphs and neglect an important property of the nodes: influence. Li, Qin, Yu, and Mao introduced a novel community model called k-influential community which is based on the concept of k-core enhanced with node influence values. In this model, we are interested not only in subgraphs that are well-connected but also have a high lower-bound on their influence. More precisely, we are interested in finding top r (with respect to influence), k-core communities. We present novel approaches that provide an impressive scalability in solving the problem for graphs of billions of edges using only a consumer-grade machine. / Graduate

We propose a large neighborhood search (LNS) algorithm to solve the periodic location routing problem (PLRP). The PLRP combines location and routing decisions over a planning horizon in which customers require visits according to a given frequency and the specific visit days can be chosen. We use parallelization strategies that can exploit the availability of multiple processors. The computational results show that the algorithms obtain better results than previous solution methods on a set of standard benchmark instances from the literature.

No description available.

Experiments were conducted to test the relative importance of the two pairs of wing and the flight musculature of Oncopeltus fasciatus. Further, the postembryonic development of this musculature was investigated. It is shown that flight is impossible with only the hind-wings present. The fore-wings are the major propulsive organs, with the hind-wings providing only a part of the lift component. The hind-wings are operated by the mesothoracic musculature acting through a hook mechanism which joins the two pairs of wings together. The development of the mesothoracic muscles is shown to be in two stages; the first involves the degeneration of the original muscle fibres present in the first instar insect, the second involves the aggregation of myoblasts to form fibres which mature by about the third day after the moult into the adult stage. Some evolutionary comments are offered on how the developmental processes described in this thesis, compare with those previously described in other insect orders. / Science, Faculty of / Zoology, Department of / Graduate

Les systèmes multi-agents sont constitués d'entités autonomes qui interagissent avec leur environnement pour résoudre un objectif collectif. Si l'on souhaite modéliser des systèmes contenant des millions d'agents, une puissance de calcul et de stockage importante devient nécessaire. Pour atteindre de telles simulations large échelle, distribuer le simulateur sur un réseau de machines est nécessaire, mais il faut prendre en compte quelques aspects. Le premier aspect se concentre sur deux types de répartition de la charge de calcul : la première basée sur l'environnement, la deuxième basée sur les agents. Nous évaluons les performances de ces répartitions en les confrontant à des applications dont les dynamiques de déplacement sont très différentes, ce qui nous permet d'identifier plusieurs critères devant être pris en compte pour garantir des gains de performance lors de la distribution de simulations d'agents situés. Le second aspect de notre travail étudie la synchronisation. En effet, à notre connaissance, tous les simulateurs existants fonctionnent sur la base d'une synchronisation forte entre les machines, ce qui garantit la causalité temporelle et la cohérence des calculs. Dans cette thèse, nous remettons en cause cette hypothèse en étudiant la relaxation de la contrainte de synchronisation. Nous proposons deux politiques de synchronisation : la synchronisation forte classique et une forme de synchronisation reposant sur une fenêtre de temps bornée entre la machine la plus lente et la machine la plus rapide. Des applications de natures différentes sont exécutées avec ces différents mécanismes de synchronisation. / This thesis aims to design a distributed large scale MAS simulation. When the number of agents reaches several millions, it is necessary to distribute MAS simulation. However, this can raise some issues: agents allocation, interactions from different machines, time management, etc. When we distribute MAS simulation on different machines, agents must be separated between these machines and should still be able to produce their normal behaviours. Our distribution is able to cover all agents' perceptions during the simulation and allow all agents to interact normally. Moreover, with large-scale simulations the main observations are done on the macroscopic level. In this thesis, we study two main aspects to distribute large-scale simulations. The first aspect is the efficient strategy that can be used to distribute MAS concepts (agents and environment). We propose two efficient distribution approaches: agents distribution and environment distribution. The second aspect is the relaxation of synchronization constraints in order to speed up the execution of large-scale simulations. Relaxing this constraint can induce incoherent interactions, which do not exist in a synchronized context. But, in some applications that can not affect the macroscopic level. Our experiments on different categories of MAS applications show that some applications can be distributed efficiently in one distribution approach more than the other. In addition, we have studied the impact of incoherent iterations on the emerging behaviour of different applications, and we have evidenced situations in which unsynchronized simulations still produced the expected macroscopic behaviour.

This international workshop entitled “How to integrate geochemistry at affordable costs into reac-tive transport for large-scale systems” was organized by the Institute of Resource Ecology of the Helmholtz-Zentrum Dresden Rossendorf in Feb-ruary 2020. A mechanistic understanding and building on that an appropriate modelling of geochemical processes is essential for reliably predicting contaminant transport in groundwater systems, but also in many other cases where migration of hazardous substances is expected and consequently has to be assessed and limited. In case of already present contaminations, such modelling may help to quantify the threads and to support the development and application of suitable remediation measures. Typical application areas are nuclear waste disposal, environmental remediation, mining and milling, carbon capture & storage, or geothermal energy production. Experts from these fields were brought together to discuss large-scale reactive transport modelling (RTM) because the scales covered by such pre-dictions may reach up to one million year and dozens of kilometers. Full-fledged incorporation of geochemical processes, e.g. sorption, precipitation, or redox reactions (to name just a few important basic processes) will thus create inacceptable long computing times. As an effective way to integrate geochemistry at affordable costs into RTM different geochemical concepts (e.g. multidimensional look-up tables, surrogate functions, machine learning, utilization of uncertainty and sensitivity analysis etc.) exist and were extensively discussed throughout the workshop. During the 3-day program of the workshop keynote and regular lectures from experts in the field, a poster session, and a radio lab tour had been offered. In total, 40 scientists from 28 re-search institutes and 8 countries participated.

In this study a new approach to active-element pattern analysis, for large phased array antennas, was created using Floquet's theorem. The classic approach to finding active-element patterns uses a full array simulation that can become slow and produce patterns that are specific to certain elements in the array, though basically identical away from the array edge. Instead of producing specific active-element patterns an average active-element pattern could be created and then applied that to the array. The average active-element pattern can be used for every element in the array with a small margin of error. Using Floquet's theorem reduces any differences between elements in the array and gives the most accurate active-element pattern within a reasonable time constraint. Floquet average active-element patterns are computed by using an infinite array and a summation is done for the far-field radiation values of a finite array based on the number of elements using typical pattern multiplication techniques. Therefore, accuracy of the Floquet element approach is excellent for arrays on the size of hundreds to thousands of elements. An active-element pattern is determined by scanning the array and taking the far-field radiation value at each beam scan-angle. Each beam scan-angle value is a summation of the element radiation patterns in that specific direction. These beam scan-angle values are then reduced by the number of elements in the array to form a radiation pattern. This radiation pattern is the average active-element pattern. / Master of Science

No description available.

Solving large-scale linear systems of equations (LSEs) is one of the most common and fundamental problems in big data. But such problems are often too expensive to solve for resource-limited users. Cloud computing has been proposed as an efficient and costeffective way of solving such tasks. Nevertheless, one critical concern in cloud computing is data privacy. Many previous works on secure outsourcing of LSEs have high computational complexity and share a common serious problem, i.e., a huge number of external memory I/O operations, which may render those outsourcing schemes impractical. We develop a practical secure outsourcing algorithm for solving large-scale LSEs, which has both low computational complexity and low memory I/O complexity and can protect clients privacy well. We implement our algorithm on a real-world cloud server and a laptop. We find that the proposed algorithm offers significant time savings for the client (up to 65%) compared to previous algorithms.

Atmospheric boundary layer flows around wind turbines distributed in a large wind farm can be examined by the use of large eddy simulation (LES), which is based on the assumption that large eddies in the flow are anisotropic and depend on the mean flow and the configuration geometry, while smaller eddies are isotropic and homogeneous, and can be modeled via subgrid scale models. In this thesis, a pseudo-spectral LES code with inflow conditions imposed through a precursor concurrent simulation is utilized to model the flow around a single wind turbine or a large wind farm operating in thermally-stratified conditions. The effect of the wind turbines on humidity is monitored through an additional scalar convection equation. It is found that on average, the effect of an individual wind turbine on the humidity is less than 1%, while the effect of the wind farm on humidity can reach 1-2% in the cumulative wakes.