Varying Resource Requirements Research Articles

Dynamic UAV Deployment Scheme Based on Edge Computing for Forest Fire Scenarios.

This study investigates the dynamic deployment of unmanned aerial vehicles (UAVs) using edge computing in a forest fire scenario. We consider the dynamically changing characteristics of forest fires and the corresponding varying resource requirements. Based on this, this paper models a two-timescale UAV dynamic deployment scheme by considering the dynamic changes in the number and position of UAVs. In the slow timescale, we use a gate recurrent unit (GRU) to predict the number of future users and determine the number of UAVs based on the resource requirements. UAVs with low energy are replaced accordingly. In the fast timescale, a deep-reinforcement-learning-based UAV position deployment algorithm is designed to enable the low-latency processing of computational tasks by adjusting the UAV positions in real time to meet the ground devices' computational demands. The simulation results demonstrate that the proposed scheme achieves better prediction accuracy. The number and position of UAVs can be adapted to resource demand changes and reduce task execution delays.

A von-Neumann-like photonic processor and its application in studying quantum signature of chaos

Photonic quantum computation plays an important role and offers unique advantages. Two decades after the milestone work of Knill-Laflamme-Milburn, various architectures of photonic processors have been proposed, and quantum advantage over classical computers has also been demonstrated. It is now the opportune time to apply this technology to real-world applications. However, at current technology level, this aim is restricted by either programmability in bulk optics or loss in integrated optics for the existing architectures of processors, for which the resource cost is also a problem. Here we present a von-Neumann-like architecture based on temporal-mode encoding and looped structure on table, which is capable of multimode-universal programmability, resource-efficiency, phase-stability and software-scalability. In order to illustrate these merits, we execute two different programs with varying resource requirements on the same processor, to investigate quantum signature of chaos from two aspects: the signature behaviors exhibited in phase space (13 modes), and the Fermi golden rule which has not been experimentally studied in quantitative way before (26 modes). The maximal program contains an optical interferometer network with 1694 freely-adjustable phases. Considering current state-of-the-art, our architecture stands as the most promising candidate for real-world applications.

Concept recognition as a machine translation problem

BackgroundAutomated assignment of specific ontology concepts to mentions in text is a critical task in biomedical natural language processing, and the subject of many open shared tasks. Although the current state of the art involves the use of neural network language models as a post-processing step, the very large number of ontology classes to be recognized and the limited amount of gold-standard training data has impeded the creation of end-to-end systems based entirely on machine learning. Recently, Hailu et al. recast the concept recognition problem as a type of machine translation and demonstrated that sequence-to-sequence machine learning models have the potential to outperform multi-class classification approaches.MethodsWe systematically characterize the factors that contribute to the accuracy and efficiency of several approaches to sequence-to-sequence machine learning through extensive studies of alternative methods and hyperparameter selections. We not only identify the best-performing systems and parameters across a wide variety of ontologies but also provide insights into the widely varying resource requirements and hyperparameter robustness of alternative approaches. Analysis of the strengths and weaknesses of such systems suggest promising avenues for future improvements as well as design choices that can increase computational efficiency with small costs in performance.ResultsBidirectional encoder representations from transformers for biomedical text mining (BioBERT) for span detection along with the open-source toolkit for neural machine translation (OpenNMT) for concept normalization achieve state-of-the-art performance for most ontologies annotated in the CRAFT Corpus. This approach uses substantially fewer computational resources, including hardware, memory, and time than several alternative approaches.ConclusionsMachine translation is a promising avenue for fully machine-learning-based concept recognition that achieves state-of-the-art results on the CRAFT Corpus, evaluated via a direct comparison to previous results from the 2019 CRAFT shared task. Experiments illuminating the reasons for the surprisingly good performance of sequence-to-sequence methods targeting ontology identifiers suggest that further progress may be possible by mapping to alternative target concept representations. All code and models can be found at: https://github.com/UCDenver-ccp/Concept-Recognition-as-Translation.

DA-DRLS: Drift adaptive deep reinforcement learning based scheduling for IoT resource management

In order to fulfill the tremendous resource demand by diverse IoT applications, the large-scale resource-constrained IoT ecosystem requires a robust resource management technique. An optimal resource provisioning in IoT ecosystem deals with an efficient request-resource mapping which is difficult to achieve due to the heterogeneity and dynamicity of IoT resources and IoT requests. In this paper, we investigate the scheduling and resource allocation problem for dynamic user requests with varying resource requirements. Specifically, we formulate the complete problem as an optimization problem and try to generate an optimal policy with the objectives to minimize the overall energy consumption and to achieve a long-term user satisfaction through minimum response time. We introduce the paradigm of a deep reinforcement learning (DRL) mechanism to escalate the resource management efficiency in IoT ecosystem. To maximize the numerical performance of the entire resource management activities, our method learns to select the optimal resource allocation policy among a number of possible solutions. Moreover, the proposed approach can efficiently handle a sudden hike or fall in users' demand, which we call demand drift, through adaptive learning maintaining the optimal resource utilization. Finally, our simulation analysis illustrates the effectiveness of the proposed mechanism as it achieves substantial improvements in various factors, like reducing energy consumption and response time by at least 36.7% and 59.7% respectively and increasing average resource utilization by at least 10.4%. Our approach also attains a good convergence and a trade-off between the monitoring metrics.

Stochastic Virtual Machine Placement for Cloud Data Centers Under Resource Requirement Variations

In cloud computing environment, the optimal placement of virtual machines (VMs) onto physical servers has been of great importance to improving the resource utilization and energy efficiency of data centers. In this work, we study the VM placement problem for minimizing the total energy consumption in a data center under the uncertainty of resource requirements demanded by the VMs. Instead of using deterministic values to represent the resource requirements, as in most existing placers, we propose a stochastic placement approach in which the resource requirement variations are modeled as random variables. We further formulate the uncertainty-aware VM placement problem as a stochastic optimization model, of which the optimization objective is to minimize the total energy consumed by all physical machines (PMs). In the presence of varying resource requirements, the optimization model is subject to a probabilistic constraint on resource overflow probability on each PM (i.e., the probability of demanded CPU/memory exceeding the maximum capacity the PM can provide). To solve this stochastic optimization problem, we develop an efficient metaheuristic to seek for an optimized VM placement solution that minimizes the total energy cost while satisfying the probabilistic resource constraint. Moreover, by incorporating a solution initialization procedure and a neighborhood search strategy, we can further improve the effectiveness of the metaheuristic in solution space exploration. Extensive simulations are performed to justify the proposed approach, in terms of both solution feasibility and energy efficiency. By taking into account the uncertainty of resource requirements, the stochastic method can achieve more energy-efficient placement solutions compared with the deterministic VM placement algorithm.

Fermion-to-qubit mappings with varying resource requirements for quantum simulation

The mapping of fermionic states onto qubit states, as well as the mapping of fermionic Hamiltonian into quantum gates enables us to simulate electronic systems with a quantum computer. Benefiting the understanding of many-body systems in chemistry and physics, quantum simulation is one of the great promises of the coming age of quantum computers. Interestingly, the minimal requirement of qubits for simulating Fermions seems to be agnostic of the actual number of particles as well as other symmetries. This leads to qubit requirements that are well above the minimal requirements as suggested by combinatorial considerations. In this work, we develop methods that allow us to trade-off qubit requirements against the complexity of the resulting quantum circuit. We first show that any classical code used to map the state of a fermionic Fock space to qubits gives rise to a mapping of fermionic models to quantum gates. As an illustrative example, we present a mapping based on a nonlinear classical error correcting code, which leads to significant qubit savings albeit at the expense of additional quantum gates. We proceed to use this framework to present a number of simpler mappings that lead to qubit savings with a more modest increase in gate difficulty. We discuss the role of symmetries such as particle conservation, and savings that could be obtained if an experimental platform could easily realize multi-controlled gates.

Cluster-Aware Virtual Machine Collaborative Migration in Media Cloud

Media cloud has become a promising paradigm for deploying large-scale streaming media applications at a reduced cost. Due to dynamic and diverse demands of users, media cloud presents two crucial characteristics: high resource consumption and dynamic traffic among media servers. Consequently, Virtual Machine (VM) migration in media cloud is highly required to suit varying resource requirements and the dynamic traffic patterns. Moreover, migration of such bandwidth-intensive media applications in media cloud needs cautious handling, especially for the internal traffic of Data Center Networks (DCN). However, existing media cloud resource management schemes or traffic-aware VM deployment approaches are insufficient for media cloud, ignoring the characteristics of either cloud infrastructure or media streaming requirements. In this paper, we propose a cluster-aware VM collaborative migration scheme for media cloud, tightly integrating clustering, placement, and dynamic migration process. The scheme employs a clustering algorithm and a placement algorithm to obtain ideal migration strategies for newly perceived media server clusters, and a migration algorithm to effectively accomplish the migration process of media servers. Evaluation results demonstrate that our scheme can effectively migrate virtual media servers in media cloud, while reducing the total internal traffic in DCN under the resource consumption constraints of media streaming applications.

An Optimization of Energy-Efficiency in Machining Manufacturing Systems Based on a Framework of Multi-Mode RCPSP

It has become important to consider energy-efficient optimization not only in a process design but also in the operations of manufacturing systems to promote sustainable and green manufacturing. This paper extends authors’ previous work to a more practical situation to demonstrate the applicability of the proposed framework of energy-efficient manufacturing operations based on a resource-constrained project scheduling problem (RCPSP). Both have varying resource requirements and multi processing modes, which can produce a suitable energy-load profiles for complete manufacturing systems. This study proposes a mathematical model for producing optimal energy-load profiles, and based on these profiles, each given operation is allocated to a machine tool with a specific processing mode. A processing mode refers to machining conditions for the corresponding operation, conditions that provide a predictive processing time and estimated electrical energy consumption. Through some cutting experiments on aluminum alloy performed on a three-axis machining center, we provide several possible processing modes for workpieces (operations), and we generate energy-load profiles by applying multi start local searches. We then discuss the applicability and capability of the energy-load profiles as an energy-aware production control.

Energy efficiency of dynamic management of virtual cluster with heterogeneous hardware

Cloud computing is an essential part of today’s computing world. Continuously increasing amount of computation with varying resource requirements is placed in large data centers. The variation among computing tasks, both in their resource requirements and time of processing, makes it possible to optimize the usage of physical hardware by applying cloud technologies. In this work, we develop a prototype system for load-based management of virtual machines in an OpenStack computing cluster. Our prototype is based on an idea of ‘packing’ idle virtual machines into special park servers optimized for this purpose. We evaluate the method by running real high-energy physics analysis software in an OpenStack test cluster and by simulating the same principle using the Cloudsim simulator software. The results show a clear improvement, 9–48 % , in the total energy efficiency when using our method together with resource overbooking and heterogeneous hardware.

Adaptive offloading for infotainment systems

Infotainment applications in vehicles are currently supported both by the in-vehicle platform, as well as by user's smart devices, such as smartphones and tablets. More and more the user expects that there is a continuous service of applications inside or outside of the vehicle, provided in any of these devices (a simple but common example is hands-free mobile phone calls provided by the vehicle platform). With the increasing complexity of 'apps', it is necessary to support increasing levels of Quality of Service (QoS), with varying resource requirements. Users may want to start listening to music in the smartphone, or video in the tablet, being this application transparently 'moved' into the vehicle when it is started. This paper presents an adaptable offloading mechanism, following a service-oriented architecture pattern, which takes into account the QoS requirements of the applications being executed when making decisions.

PRISM: Fine-Grained Resource-Aware Scheduling for MapReduce

MapReduce has become a popular model for data-intensive computation in recent years. By breaking down each job into small map and reduce tasks and executing them in parallel across a large number of machines, MapReduce can significantly reduce the running time of data-intensive jobs. However, despite recent efforts toward designing resource-efficient MapReduce schedulers, existing solutions that focus on scheduling at the task-level still offer sub-optimal job performance. This is because tasks can have highly varying resource requirements during their lifetime, which makes it difficult for task-level schedulers to effectively utilize available resources to reduce job execution time. To address this limitation, we introduce PRISM, a fine-grained resource-aware MapReduce scheduler that divides tasks into phases, where each phase has a constant resource usage profile, and performs scheduling at the phase level. We first demonstrate the importance of phase-level scheduling by showing the resource usage variability within the lifetime of a task using a wide-range of MapReduce jobs. We then present a phase-level scheduling algorithm that improves execution parallelism and resource utilization without introducing stragglers. In a 10-node Hadoop cluster running standard benchmarks, PRISM offers high resource utilization and provides 1.3× improvement in job running time compared to the current Hadoop schedulers.

Architecture for adaptive resource assignment to virtualized mixed-criticality real-time systems

System virtualization is a powerful approach for the creation of integrated systems, which meet the high functionality and reliability requirements of complex embedded applications. It is in particular well-suited for mixed-criticality systems, since the often applied pessimistic manner of critical system engineering leads to heavily under-utilized resources. Existing static resource management approaches for virtualized systems are inappropriate for the dynamically varying resource requirements of upcoming adaptive systems. In this paper, we propose a dynamic resource management protocol for system virtualization that factors criticality levels in and allows the addition of subsystems at runtime. The two-level architecture offers flexibility across virtual machine borders and has the potential to improve the resource utilization. In addition, it provides the capability to adapt at runtime according to defects or changes of the environment.

Basics of virtual machine migration on heterogeneous architectures for self-optimizing mechatronic systems

The combination of virtualization and heterogeneous multi-processor architectures supports the development of efficient platforms for self-optimizing mechatronic systems, which are characterized by varying resource requirements. Virtualization addresses dependability issues and adds runtime flexibility to provide an appropriate resource management. Real-time requirements have to be met and existing real-time virtualization solutions are characterized by a static mapping of virtual machines to processors and do not use the full potential of heterogeneous architectures. The sketched architecture applies migration and emulation to realize a dynamic assignment of virtual machines to processors. This work identifies the necessary conditions for migration and the degree of communication between hypervisor and operating system that is indispensable for migration decisions. System virtualization is an approach of coarse-grained system consolidation, for which reason implementation issues and possibilities to reduce the high overhead are discussed. The implementation of a real-time capable virtual machine migration requires paravirtualization, that is to say a modification of the guest operating systems. The need to modify the guest operating system is outweighed by the advantages in terms of flexibility of an explicit communication and the resulting cooperation of hypervisor and guest operating system.

Modeling and exploiting spatial locality trade-offs in wavelet-based applications under varying resource requirements

Future dynamic applications will require new mapping strategies to deliver power-efficient performance. Fully static design-time mappings will not be able to optimally address the unpredictably varying application characteristics and system resource requirements. Instead, the platforms will not only need to be programmable in terms of instruction set processors, but also at least partial reconfigurability will be required, while the applications themselves will need to exploit this increased freedom at runtime to adapt to the dynamism. In this context, it is important for applications to optimally exploit the memory hierarchy under varying memory availability. This article presents an analysis of spatial locality trade-offs in wavelet-based applications, to be used in dynamic execution environments: Depending on the encountered runtime conditions, the execution switches to different memory optimized instantiations or localizations, optimally exploiting temporal and spatial locality under these conditions. This is enabled by systematic mapping guidelines, indicating how the miss-rate behavior of a localization is influenced by a specific execution condition, under which conditions a certain localization is optimal and which miss-rate gains may be obtained by switching to that localization.

Exploiting Varying Resource Requirements in Wavelet-based Applications in Dynamic Execution Environments

In the context of future dynamic applications, systems will exhibit unpredictably varying platform resource requirements. To deal with this, they will not only need to be programmable in terms of instruction set processors, but also at least partial reconfigurability will be required. In this context, it is important for applications to optimally exploit the memory hierarchy under varying memory availability. This article presents a mapping strategy for wavelet-based applications: depending on the encountered conditions, it switches to different memory optimized instantations or localizations, permitting up to 51% energy gains in memory accesses. Systematic and parameterized mapping guidelines indicate which localization should be selected when, for varying algorithmic wavelet parameters. The results have been formalized and generalized to be applicable to more general wavelet-based applications.

A LP-based Admission Control Using Artificial Neural Networks for Integrated Services in Mobile Networks

In mobile networks the traffic fluctuation is unpredictable due to mobility and varying resource requirements of multimedia applications. Hence it is essential to maintain the traffic within the network capacity to provide the service guarantees to running applications. This paper proposes an Admission Control (AC) scheme in a single mobile cellular environment supporting real-time and non-real-time application traffic. In the case of a real-time and non-real-time multimedia applications, each application has its own distinct range of acceptable Quality of Service (QoS) requirements (e.g., packet loss, delay, jitter, etc.). The network provides the service by maintaining the application specified QoS range. We propose a Linear Programming Resource Reduction (LP-RR) principle for admission control by maintaining QoS guarantees to existing applications and to increase the percentage of admission to real-time and non-real-time applications. Artificial Neural Networks (ANNs) are used to solve linear programming problem, which facilitates an on-line admission control decision in the practical systems. The simulation results demonstrate that the proposed AC scheme performs well in terms of admitted applications and maintains lower percentage of rejection to hand-off and new applications of different traffic classes. The suggested principle also shown that it is appropriate for the fair resource allocation with improved resource utilization.

Dynamic resource brokering for multi-user query execution

We propose a new framework for resource allocation based on concepts from microeconomics. Specifically, we address the difficult problem of managing resources in a multiple-query environment composed of queries with widely varying resource requirements. The central element of the framework is a resource broker that realizes a profit by "selling" resources to competing operators using a performance-based "currency." The guiding principle for brokering resources is profit maximization. In other words, since the currency is derived from the performance objective, the broker can achieve the best performance by making the scheduling and resource allocation decisions that maximize profit. Moreover, the broker employs dynamic techniques and adapts by changing previous allocation decisions while queries are executing. In a first validation study of the framework, we developed a prototype broker that manages memory and disk bandwidth for a multi-user query workload. The performance objective for the prototype broker is to minimize slowdown with the constraint of fairness. Slowdown measures how much higher the response time is in a multi-user environment than a single-user environment, and fairness measures how even is the degradation in response time among all queries as the system load increases, Our simulation results show the viability of the broker framework and the effectiveness of our query admission and resource allocation policies for multi-user workloads.