Dynamic SLA management with forecasting using multi-objective optimization

Cost-efficient operation while satisfying performance and availability guarantees in Service Level Agreements (SLAs) is a challenge for Cloud Computing, as these are potentially conflicting objectives. We present a framework for SLA management based on multi-objective optimization. The framework features a forecasting model for determining the best virtual machine-to-host allocation given the need to minimize SLA violations, energy consumption and resource wasting. A comprehensive SLA management solution is proposed that uses event processing for monitoring and enables dynamic provisioning of virtual machines onto the physical infrastructure. We validated our implementation against serveral standard heuristics and were able to show that our approach is significantly better.

Adapting agricultural land management to climate change: a regional multi-objective optimization approach

In several regions of the world, climate change is expected to have severe impacts on agricultural systems. Changes in land management are one way to adapt to future climatic conditions, including land-use changes and local adjustments of agricultural practices. In previous studies, options for adaptation have mostly been explored by testing alternative scenarios. Systematic explorations of land management possibilities using optimization approaches were so far mainly restricted to studies of land and resource management under constant climatic conditions. In this study, we bridge this gap and exploit the benefits of multi-objective regional optimization for identifying optimum land management adaptations to climate change. We design a multi-objective optimization routine that integrates a generic crop model and considers two climate scenarios for 2050 in a meso-scale catchment on the Swiss Central Plateau with already limited water resources. The results indicate that adaptation will be necessary in the study area to cope with a decrease in productivity by 0–10 %, an increase in soil loss by 25–35 %, and an increase in N-leaching by 30–45 %. Adaptation options identified here exhibit conflicts between productivity and environmental goals, but compromises are possible. Necessary management changes include (i) adjustments of crop shares, i.e. increasing the proportion of early harvested winter cereals at the expense of irrigated spring crops, (ii) widespread use of reduced tillage, (iii) allocation of irrigated areas to soils with low water-retention capacity at lower elevations, and (iv) conversion of some pre-alpine grasslands to croplands.

A Genetic Algorithm to associate optical measurements and estimate orbits of geocentric objects

Cataloging geocentric objects can be put in the framework of Multiple Target Tracking (MTT). Current work tends to focus on the S = 2 MTT problem because of its favorable computational complexity of O(n²). The MTT problem becomes NP-hard for a dimension of S˃3. The challenge is to find an approximation to the solution within a reasonable computation time. To effciently approximate this solution a Genetic Algorithm is used. The algorithm is applied to a simulated test case. These results represent the first steps towards a method that can treat the S˃3 problem effciently and with minimal manual intervention.

A study of the time complexity of an Elitist Genetic Algorithm used for associating optical observations of MEO and GEO Space Debris

Currently several thousands of objects are being tracked in the MEO and GEO regions through optical means. The problem faced in this framework is that of Multiple Target Tracking (MTT). In this context both the correct associations among the observations, and the orbits of the objects have to be determined. The complexity of the MTT problem is defined by its dimension S. Where S stands for the number of ’fences’ used in the problem, each fence consists of a set of observations that all originate from dierent targets. For a dimension of S ˃ the MTT problem becomes NP-hard. As of now no algorithm exists that can solve an NP-hard problem in an optimal manner within a reasonable (polynomial) computation time. However, there are algorithms that can approximate the solution with a realistic computational e ort. To this end an Elitist Genetic Algorithm is implemented to approximately solve the S ˃ MTT problem in an e cient manner. Its complexity is studied and it is found that an approximate solution can be obtained in a polynomial time. With the advent of improved sensors and a heightened interest in the problem of space debris, it is expected that the number of tracked objects will grow by an order of magnitude in the near future. This research aims to provide a method that can treat the correlation and orbit determination problems simultaneously, and is able to e ciently process large data sets with minimal manual intervention.

Service level agreements-driven management of distributed applications in cloud computing environments

Advancements in cloud computing have enabled the proliferation of distributed applications, which require management and control of multiple services. However, without an efficient mechanism for scaling services in response to changing workload conditions, such as number of connected users, application performance might suffer, leading to violations of Service Level Agreements (SLA) and possible inefficient use of hardware resources. Combining dynamic application requirements with the increased use of virtualised computing resources creates a challenging resource Management context for application and cloud-infrastructure owners. In such complex environments, business entities use SLAs as a means for specifying quantitative and qualitative requirements of services. There are several challenges in running distributed enterprise applications in cloud environments, ranging from the instantiation of service VMs in the correct order using an adequate quantity of computing resources, to adapting the number of running services in response to varying external loads, such as number of users. The application owner is interested in finding the optimum amount of computing and network resources to use for ensuring that the performance requirements of all her/his applications are met. She/he is also interested in appropriately scaling the distributed services so that application performance guarantees are maintained even under dynamic workload conditions. Similarly, the infrastructure Providers are interested in optimally provisioning the virtual resources onto the available physical infrastructure so that her/his operational costs are minimized, while maximizing the performance of tenants’ applications. Motivated by the complexities associated with the management and scaling of distributed applications, while satisfying multiple objectives (related to both consumers and providers of cloud resources), this thesis proposes a cloud resource management platform able to dynamically provision and coordinate the various lifecycle actions on both virtual and physical cloud resources using semantically enriched SLAs. The system focuses on dynamic sizing (scaling) of virtual infrastructures composed of virtual machines (VM) bounded application services. We describe several algorithms for adapting the number of VMs allocated to the distributed application in response to changing workload conditions, based on SLA-defined performance guarantees. We also present a framework for dynamic composition of scaling rules for distributed service, which used benchmark-generated application Monitoring traces. We show how these scaling rules can be combined and included into semantic SLAs for controlling allocation of services. We also provide a detailed description of the multi-objective infrastructure resource allocation problem and various approaches to satisfying this problem. We present a resource management system based on a genetic algorithm, which performs allocation of virtual resources, while considering the optimization of multiple criteria. We prove that our approach significantly outperforms reactive VM-scaling algorithms as well as heuristic-based VM-allocation approaches.

SOP: Parallel surrogate global optimization with Pareto center selection for computationally expensive single objective problems

This paper presents a parallel surrogate-based global optimization method for computationally expensive objective functions that is more effective for larger numbers of processors. To reach this goal, we integrated concepts from multi-objective optimization and tabu search into, single objective, surrogate optimization. Our proposed derivative-free algorithm, called SOP, uses non-dominated sorting of points for which the expensive function has been previously evaluated. The two objectives are the expensive function value of the point and the minimum distance of the point to previously evaluated points. Based on the results of non-dominated sorting, P points from the sorted fronts are selected as centers from which many candidate points are generated by random perturbations. Based on surrogate approximation, the best candidate point is subsequently selected for expensive evaluation for each of the P centers, with simultaneous computation on P processors. Centers that previously did not generate good solutions are tabu with a given tenure. We show almost sure convergence of this algorithm under some conditions. The performance of SOP is compared with two RBF based methods. The test results show that SOP is an efficient method that can reduce time required to find a good near optimal solution. In a number of cases the efficiency of SOP is so good that SOP with 8 processors found an accurate answer in less wall-clock time than the other algorithms did with 32 processors.

Graph-Based Multi-Surface Segmentation of OCT Data Using Trained Hard and Soft Constraints

Optical coherence tomography (OCT) is a well-established image modality in ophthalmology and used daily in the clinic. Automatic evaluation of such datasets requires an accurate segmentation of the retinal cell layers. However, due to the naturally low signal to noise ratio and the resulting bad image quality, this task remains challenging. We propose an automatic graph-based multi-surface segmentation algorithm that internally uses soft constraints to add prior information from a learned model. This improves the accuracy of the segmentation and increase the robustness to noise. Furthermore, we show that the graph size can be greatly reduced by applying a smart segmentation scheme. This allows the segmentation to be computed in seconds instead of minutes, without deteriorating the segmentation accuracy, making it ideal for a clinical setup. An extensive evaluation on 20 OCT datasets of healthy eyes was performed and showed a mean unsigned segmentation error of 3.05 ±0.54 μm over all datasets when compared to the average observer, which is lower than the inter-observer variability. Similar performance was measured for the task of drusen segmentation, demonstrating the usefulness of using soft constraints as a tool to deal with pathologies.

Evaluation of texture features in hepatic tissue characterization from non-enhanced CT images

Aim of this paper is to evaluate the diagnostic contribution of various types of texture features in discrimination of hepatic tissue in abdominal non-enhanced Computed Tomography (CT) images. Regions of Interest (ROIs) corresponding to the classes: normal liver, cyst, hemangioma, and hepatocellular carcinoma were drawn by an experienced radiologist. For each ROI, five distinct sets of texture features are extracted using First Order Statistics (FOS), Spatial Gray Level Dependence Matrix (SGLDM), Gray Level Difference Method (GLDM), Laws' Texture Energy Measures (TEM), and Fractal Dimension Measurements (FDM). In order to evaluate the ability of the texture features to discriminate the various types of hepatic tissue, each set of texture features, or its reduced version after genetic algorithm based feature selection, was fed to a feed-forward Neural Network (NN) classifier. For each NN, the area under Receiver Operating Characteristic (ROC) curves (Az) was calculated for all one-vs-all discriminations of hepatic tissue. Additionally, the total Az for the multi-class discrimination task was estimated. The results show that features derived from FOS perform better than other texture features (total Az: 0.802+/-0.083) in the discrimination of hepatic tissue.

Service Level Agreements-Driven Management of Distributed Applications in Cloud Computing Environments

Cloud Computing enables provisioning and distribution of highly scalable services in a reliable, on-demand and sustainable manner. However, objectives of managing enterprise distributed applications in cloud environments under Service Level Agreement (SLA) constraints lead to challenges for maintaining optimal resource control. Furthermore, conflicting objectives in management of cloud infrastructure and distributed applications might lead to violations of SLAs and inefficient use of hardware and software resources. This dissertation focusses on how SLAs can be used as an input to the cloud management system, increasing the efficiency of allocating resources, as well as that of infrastructure scaling. First, we present an extended SLA semantic model for modelling complex service-dependencies in distributed applications, and for enabling automated cloud infrastructure management operations. Second, we describe a multi-objective VM allocation algorithm for optimised resource allocation in infrastructure clouds. Third, we describe a method of discovering relations between the performance indicators of services belonging to distributed applications and then using these relations for building scaling rules that a CMS can use for automated management of VMs. Fourth, we introduce two novel VM-scaling algorithms, which optimally scale systems composed of VMs, based on given SLA performance constraints. All presented research works were implemented and tested using enterprise distributed applications.

A computer-aided diagnostic system to characterize CT focal liver lesions: design and optimization of a neural network classifier

In this paper, a computer-aided diagnostic (CAD) system for the classification of hepatic lesions from computed tomography (CT) images is presented. Regions of interest (ROIs) taken from nonenhanced CT images of normal liver, hepatic cysts, hemangiomas, and hepatocellular carcinomas have been used as input to the system. The proposed system consists of two modules: the feature extraction and the classification modules. The feature extraction module calculates the average gray level and 48 texture characteristics, which are derived from the spatial gray-level co-occurrence matrices, obtained from the ROIs. The classifier module consists of three sequentially placed feed-forward neural networks (NNs). The first NN classifies into normal or pathological liver regions. The pathological liver regions are characterized by the second NN as cyst or "other disease." The third NN classifies "other disease" into hemangioma or hepatocellular carcinoma. Three feature selection techniques have been applied to each individual NN: the sequential forward selection, the sequential floating forward selection, and a genetic algorithm for feature selection. The comparative study of the above dimensionality reduction methods shows that genetic algorithms result in lower dimension feature vectors and improved classification performance.

Tropopause level Rossby wave breaking in the Northern Hemisphere: a feature-based validation of the ECHAM5-HAM climate model

Breaking synoptic-scale Rossby waves (RWB) at the tropopause level are central to the daily weather evolution in the extratropics and the subtropics. RWB leads to pronounced meridional transport of heat, moisture, momentum, and chemical constituents. RWB events are manifest as elongated and narrow structures in the tropopause-level potential vorticity (PV) field. A feature-based validation approach is used to assess the representation of Northern Hemisphere RWB in present-day climate simulations carried out with the ECHAM5-HAM climate model at three different resolutions (T42L19, T63L31, and T106L31) against the ERA-40 reanalysis data set. An objective identification algorithm extracts RWB events from the isentropic PV field and allows quantifying the frequency of occurrence of RWB. The biases in the frequency of RWB are then compared to biases in the time mean tropopause-level jet wind speeds. The ECHAM5-HAM model captures the location of the RWB frequency maxima in the Northern Hemisphere at all three resolutions. However, at coarse resolution (T42L19) the overall frequency of RWB, i.e. the frequency averaged over all seasons and the entire hemisphere, is underestimated by 28%.The higher-resolution simulations capture the overall frequency of RWB much better, with a minor difference between T63L31 and T106L31 (frequency errors of −3.5 and 6%, respectively). The number of large-size RWB events is significantly underestimated by the T42L19 experiment and well represented in the T106L31 simulation. On the local scale, however, significant differences to ERA-40 are found in the higher-resolution simulations. These differences are regionally confined and vary with the season. The most striking difference between T106L31 and ERA-40 is that ECHAM5-HAM overestimates the frequency of RWB in the subtropical Atlantic in all seasons except for spring. This bias maximum is accompanied by an equatorward extension of the subtropical westerlies.

Analyzing diffuse scattering with supercomputers

Two new approaches to quantitatively analyze diffuse diffraction intensities from faulted layer stacking are reported. The parameters of a probability-based growth model are determined with two iterative global optimization methods: a genetic algorithm (GA) and particle swarm optimization (PSO). The results are compared with those from a third global optimization method, a differential evolution (DE) algorithm [Storn & Price (1997). J. Global Optim. 11, 341–359]. The algorithm efficiencies in the early and late stages of iteration are compared. The accuracy of the optimized parameters improves with increasing size of the simulated crystal volume. The wall clock time for computing quite large crystal volumes can be kept within reasonable limits by the parallel calculation of many crystals (clones) generated for each model parameter set on a super- or grid computer. The faulted layer stacking in single crystals of trigonal three-pointedstar- shaped tris(bicylco[2.1.1]hexeno)benzene molecules serves as an example for the numerical computations. Based on numerical values of seven model parameters (reference parameters), nearly noise-free reference intensities of 14 diffuse streaks were simulated from 1280 clones, each consisting of 96 000 layers (reference crystal). The parameters derived from the reference intensities with GA, PSO and DE were compared with the original reference parameters as a function of the simulated total crystal volume. The statistical distribution of structural motifs in the simulated crystals is in good agreement with that in the reference crystal. The results found with the growth model for layer stacking disorder are applicable to other disorder types and modeling techniques, Monte Carlo in particular.

Eco-Evolutionary Spatial Dynamics. Rapid Evolution and Isolation Explain Food Web Persistence

One of the current challenges in evolutionary ecology is understanding the long-term persistence of contemporary-evolving predator–prey interactions across space and time. To address this, we developed an extension of a multi-locus, multi-trait eco-evolutionary individual-based model that incorporates several interacting species in explicit landscapes. We simulated eco-evolutionary dynamics of multiple species food webs with different degrees of connectance across soil-moisture islands. A broad set of parameter combinations led to the local extinction of species, but some species persisted, and this was associated with (1) high connectance and omnivory and (2) ongoing evolution, due to multi-trait genetic variability of the embedded species. Furthermore, persistence was highest at intermediate island distances, likely because of a balance between predation-induced extinction (strongest at short island distances) and the coupling of island diversity by top predators, which by travelling among islands exert global top-down control of biodiversity. In the simulations with high genetic variation, we also found widespread trait evolutionary changes indicative of eco-evolutionary dynamics. We discuss how the ever-increasing computing power and high-resolution data availability will soon allow researchers to start bridging the in vivo–in silico gap.

Quantifying uncertainty on Pareto fronts with Gaussian process conditional simulations

Multi-objective optimization algorithms aim at finding Pareto-optimal solutions. Recovering Pareto fronts or Pareto sets from a limited number of function evaluations are challenging problems. A popular approach in the case of expensive-to-evaluate functions is to appeal to metamodels. Kriging has been shown efficient as a base for sequential multi-objective optimization, notably through infill sampling criteria balancing exploitation and exploration such as the Expected Hypervolume Improvement. Here we consider Kriging metamodels not only for selecting new points, but as a tool for estimating the whole Pareto front and quantifying how much uncertainty remains on it at any stage of Kriging-based multi-objective optimization algorithms. Our approach relies on the Gaussian process interpretation of Kriging, and bases upon conditional simulations. Using concepts from random set theory, we propose to adapt the Vorob’ev expectation and deviation to capture the variability of the set of non-dominated points. Numerical experiments illustrate the potential of the proposed workflow, and it is shown on examples how Gaussian process simulations and the estimated Vorob’ev deviation can be used to monitor the ability of Kriging-based multi-objective optimization algorithms to accurately learn the Pareto front.