Abstract Graph-structured data are integral to applications like social networks, biological systems, cybersecurity, and fraud detection. Outsourcing these data to public clouds offers scalability but raises privacy concerns, as encryption is required before outsourcing, making traditional graph similarity search and access control challenging. This paper presents a novel solution for privacy-preserving full graph similarity search with fine-grained access control in cloud environments. To the best of our knowledge, this is the first work to integrate privacy-preserving graph similarity search in a multi-user/multi-query setting with attribute-based access control (ABAC). This enables scalable and secure access in realistic, collaborative environments. The graph owner leverages the neural Graph2vec model to create feature indexes for encrypted graph data. Simultaneously, a secure transfer learning mechanism enables graph users to generate query feature indexes in the same latent space, ensuring privacy while accurately capturing the user’s query intent. ABAC is employed to enforce flexible, fine-grained access policies. We conduct a formal security analysis under known-ciphertext and known-background threat models, demonstrating strong privacy guarantees. Experimental evaluations on real-world datasets show that our scheme achieves high semantic accuracy, lower search latency, and reduced storage overhead, outperforming existing approaches.

Federated Learning (FL) enables collaborative model training across various distributed devices without sharing raw data. Client failures, variable energy availability, and outages of edge servers contribute to unreliable training participation, incomplete model updates, and failures at the system level during the aggregation process. In this study, we introduce a reliability-aware workload allocation FL framework (FedRAW) aimed at improving system reliability in failure-prone edge computing systems. Our approach dynamically modifies client workloads based on their failure history and integrates a lightweight backup mechanism to maintain aggregation continuity during edge server failures by backup servers handling. Additionally, we employ Bayesian optimization to fine-tune workload parameters, achieving improved energy efficiency. Experimental results reveal that our proposed method improves model accuracy while reducing energy consumption compared to recent federated learning algorithms.

Serverless computing is a cloud computing model that abstracts server management, allowing developers to focus solely on writing code without concerns about the underlying infrastructure. This paradigm shift is transforming application development by reducing time to market, lowering costs, and enhancing scalability. In serverless computing, functions are event-driven and automatically scale in response to events such as data changes or user requests. Despite its advantages, serverless computing presents several research challenges, including managing state for ephemeral functions, mitigating cold start delays, optimizing function composition, debugging, efficient auto-scaling, resource management, and ensuring security and compliance. This special issue focused on addressing these challenges by promoting research on innovative solutions and exploring the potential of serverless computing in new application domains.

In recent years, edge computing has emerged as a promising solution for deploying IoT applications that demand minimal latency. By leveraging Function as a Service (FaaS) at the edge, it is possible to achieve efficient and scalable computing capabilities. However, implementing serverless deployment at the edge presents challenges such as auto-scaling, resource management, and mitigating cold-start delays, particularly due to the limited resources available. These challenges are even more significant in workflow-based applications, where tasks are interdependent. This article introduces a dynamic approach for executing serverless workflows at the edge, consisting of three key components: initial function placement, request scheduling, and dynamic adjustment. The initial placement leverages the Non-dominated Sorting Genetic Algorithm II (NSGA-II) to deploy function instances across edge nodes. Request scheduling, on the other hand, distributes requests among these instances using a pattern graph matching algorithm. Finally, the dynamic adjustment component periodically refines placement and scheduling strategies to adapt to changing demands, utilizing a local search technique known as simulated annealing. Evaluation results indicate that the proposed solution reduces the average makespan of workflows by up to 86% compared to state-of-the-art methods.

Performing human gaze estimation using smartphones is invaluable in human-computer interaction with various potential appli- cations, ranging from user interface enhancements to medical research. We developed three deep learning-based mobile device eye-tracking architectures for dynamic visual stimuli. This includes a combination of Convolutional Neural Networks (CNN) with two different Recurrent Neural Networks (RNN), namely Long Short Term Memory (LSTM) and Gated Recurrent Unit (GRU). Our CNN+LSTM and CNN+GRU models achieved an average Root Mean Square Error of 0.955cm and 1.091cm, respectively. The codes and models are available on https://github.com/NishNilanka/MobileEye.git .

Low Earth Orbit (LEO) satellite constellations have diverse applications, including earth observation, communication services, navigation, and positioning. These constellations have evolved into a valuable data source; however, their use in a ground station (GS) for analysis via machine learning algorithms presents challenges due to constraints on power consumption, communication bandwidth, and onboard computing capabilities. While the combination of Federated Learning (FL) and Orbital Edge Computing has been employed to address these challenges, its heavy reliance on the GS for model aggregation and edge resource limitations remains a research challenge. This article presents FedOrbit, a novel energy-efficient and decentralised FL method to optimise communication with the GS and reduce power consumption. FedOrbit utilises reinforcement learning for cluster formation, satellite visiting patterns for master satellite selection, and block minifloat arithmetic for power reduction. Extensive performance evaluation under Walker Delta-based LEO constellation configurations and different datasets reveals that FedOrbit can maintain high accuracy while significantly reduce communication demand, power consumption and training time in comparison to state-of-the-art FL approaches. The proposed technique can also reduce the training time by 5× compared with the centralised FL approaches. In addition, the utilisation of block minifloat representation as low-precision arithmetic enhanced the energy consumption by 3.5× compared with the single-precision (FP32) format.

Low Earth Orbit (LEO) satellite constellations are used for a wide range of applications including earth observation, communication services, navigation, and positioning. They have emerged as a new source of data but transferring this data to a ground station (GS) for analysis and machine learning requires extensive bandwidth and incurs high latency. Limited battery capacity, communication and computing capabilities are other factors affecting the training process. Federated Learning (FL) is being used to address these challenges, although it heavily relies on the GS for model aggregation. In this paper, we consider Orbital Edge Computing (OEC) as an architecture for LEO satellite constellations and propose an on-board Federated Learning to reduce communication with the GS. We present a novel decentralised FL algorithm, called FedOrbit, based on reinforcement learning cluster formation and satellite visiting patterns to utilise intra and inter-satellite communications for model aggregation. Extensive performance evaluation under Walker Delta-based LEO constellation configurations and different datasets including MNIST, CIFAR-10, and EuroSat revealed that FedOrbit can significantly reduce communication rounds, power consumption and training time in comparison to state-of-the-art FL approaches while maintaining a high accuracy. FedOrbit demonstrates a significant decrease in power consumption, specifically by 8.8% and 79.1% for the MNIST dataset, when compared to decentralised and centralised FL approaches, respectively. The proposed technique can also reduce the training time by 5× and 48× compared with the decentralised and centralised FL approaches, respectively.

In recent years, the landscape of computing paradigms has witnessed a gradual yet remarkable shift from monolithic computing to distributed and decentralized paradigms such as Internet of Things (IoT), Edge, Fog, Cloud, and Serverless. The frontiers of these computing technologies have been boosted by shift from manually encoded algorithms to Artificial Intelligence (AI)-driven autonomous systems for optimum and reliable management of distributed computing resources. Prior work focuses on improving existing systems using AI across a wide range of domains, such as efficient resource provisioning, application deployment, task placement, and service management. This survey reviews the evolution of data-driven AI-augmented technologies and their impact on computing systems. We demystify new techniques and draw key insights in Edge, Fog and Cloud resource management-related uses of AI methods and also look at how AI can innovate traditional applications for enhanced Quality of Service (QoS) in the presence of a continuum of resources. We present the latest trends and impact areas such as optimizing AI models that are deployed on or for computing systems. We layout a roadmap for future research directions in areas such as resource management for QoS optimization and service reliability. Finally, we discuss blue-sky ideas and envision this work as an anchor point for future research on AI-driven computing systems.

Eye-tracking has been used in various domains, including human-computer interaction, psychology, and many others. Compared to commercial eye trackers, eye tracking using off-the-shelf cameras has many advantages, such as lower cost, pervasiveness, and mobility. Quantifying human attention on the mobile device is invaluable in human-computer interaction. Like videos and mobile games, dynamic visual stimuli require higher attention than static visual stimuli such as web pages and images. This research aims to develop an accurate eye-tracking algorithm using the front-facing camera of mobile devices to identify human attention hotspots when viewing video type contents. The shortage of computational power in mobile devices becomes a challenge to obtain higher user satisfaction. Edge computing moves the processing power closer to the source of the data and reduces the latency introduced by the cloud computing. Therefore, the proposed algorithm will be extended with mobile edge computing to provide a real-time eye tracking experience for users

Eye-tracking is a technique used for determining where users are looking and how long they keep their gaze fixed on a particular location. Developments in mobile technology have made mobile applications pervasive; however, eye tracking on mobile devices is still uncommon. This paper proposes a mobile edge computing architecture for eye tracking. We evaluate four lightweight CNN models (LeNet-5, AlexNet, MobileNet, and ShuffleNet) for gaze estimation on mobile devices using a publicly available dataset called GazeCapture. In order to analyse the feasibility of different inference modes such as on-device, edge-based and cloud-based, we conduct an empirical measurement study to quantify inference time, communication time, and resource consumption in these inference modes. Our analysis indicates that while cloud-based inference provides faster predictions, the communication time between the mobile device and the cloud introduces significant latency into the application. This effectively eliminates the ability to perform real-time eye tracking via cloud inference. Furthermore, our findings show that on-device inference performance is limited by energy and memory consumption, making it unsuitable to provide a high-quality user experience. Additionally, we demonstrated that edge-based inference results in a reasonable response time, memory usage, and energy consumption for eye-tracking applications on mobile devices.

This work applies Game Theory to developing secure software. With the perspective of Game Theory, one can see secure software development as a game between software developers and software security engineers, who play this game repeatedly in processes such as agile Software Development Life Cycle (SDLC). The problem we observe is that there can be conflicts between these two players regarding who should find and fix certain software vulnerabilities. To solve this problem, our approach uses Mechanism Design in Game Theory to design games that enforce cooperation between these two players. In doing so, we identify the source of the conflicts between them by looking at the components of the software. These components may be the methods or functions in the software, or individual modules, or similar building blocks. The novelty of our work is that our mechanism constructs a game which allocates software components between these two players such that they work cooperatively while trying to maximize their own payoffs.

Deep learning is a growing scientific research trend in machine learning and artificial intelligence due to its better performance compared to other machine learning techniques. This special issue focuses on recent advances in deep learning for human health care applications such as biometrics, medical imaging, and data science. The following articles were carefully reviewed and selected for this special issue: Bhurane, Dhok, Sharma, Yuvaraj, Murugappan and Acharya, ‘Diagnosis of Parkinson's disease from electroencephalography signals using linear and self-similarity features’. In this paper, the authors propose a natural (time) domain technique for diagnosing Parkinson's disease (PD). The presented computer-aided diagnosis system can act as an assistive tool to confirm the finding of PD for the clinicians. They demonstrate that using the support vector machines (SVM) classifier, the feature ranking, and the principal component analysis technique, the proposed system can detect the PD signals automatically with maximum accuracy of 99.1% ± 0.1%. Khan, Sharif, Raza, Anjum, Saba and Shad, ‘Skin lesion segmentation and classification: A unified framework of deep neural network features fusion and selection.’ This paper addresses the problem of automated skin lesion diagnosis from dermoscopic images overcoming challenges such as hairs, irregularities, lesion shape, and irrelevant feature extraction. The authors propose a hybrid approach combining optimized colour feature of lesion segmentation improved by an existing saliency approach fused with a novel pixel-based method and deep convolutional neural network (DCNN)-based skin lesion classification. Experimental results of the proposed approach demonstrate remarkable performance on three different datasets. Sivan, Sellappa and Peter J, ‘Proximity-based cloud resource provisioning for deep learning applications in smart healthcare.’ Health professionals can use smart mobile devices to convey recordings of patients and use machine learning-based approaches to process results and get predictions through smart mobile healthcare applications. Due to the nature of deep learning techniques, learning and prediction processes are moved to the cloud. This paper proposes a proximity-based resource provisioning technique that guarantees minimal delay in obtaining inference results with a local mobile cloud system. The authors implemented a healthcare cloud-based system that outperforms the state-of-the-art methods in terms of response time, deadline meeting percentage and system utilization. Arshad, Khan, Sharif, Yasmin, Tavares, Zhang and Satapathy, ‘A multilevel paradigm for deep convolutional neural network features selection with an application to human gait recognition.’ This paper proposes an integrated framework for human gait recognition using deep neural network features fusion and fuzzy entropy controlled skewness approach for best feature selection. Pre-trained CNN models (VGG19 and AlexNet) are used, and their information is mixed by the parallel fusion approach. Remarkable results on four gait analysis datasets show that the fusion of multiple CNN frameworks improves the recognition accuracy and the selection of the best features enhances the system accuracy and even minimizes the execution time. Alizadehsani, Roshanzamir, Abdar, Beykikhoshk, Khosravi, Nahavandi, Plawiak, Tan and Acharya, ‘Hybrid genetic-discretized algorithm to handle data uncertainty in diagnosing stenosis of coronary arteries.’ The authors address the uncertainty coming from noise in the data used for automated coronary artery disease (CAD) detection. They propose a novel new feature selection algorithm for CAD prediction. Authors use the genetic algorithm to determine the hyper-parameters of the SVM kernels. The system with the proposed approach demonstrates high accuracy for the stenosis diagnosis of each main coronary artery, which can help the clinicians validate their manual stenosis diagnosis of right coronary artery (RCA), right coronary artery (RCA), left circumflex (LCX) and artery and left anterior descending (LAD) coronary arteries. The results show that discretization and assurance feature selection can significantly improve the efficiency of classification algorithms. Sampathila, Pavithra and Martis, ‘Computational approach for content-based image retrieval of K-Similar images from brain MR image database.’ The task of retrieving medical images from a large image database becomes more tedious due to variations in the size and shape of the images. This paper proposes a system for content-based medical image retrieval that are relevant to a given query image. Various features such as colour, shape, and texture are exploited using the K-nearest neighbour algorithm to find the minimum distance between query and database images. The authors focus on the application of retrieving the brain MRI images of different planes (coronal, sagittal and transverse) from a dataset of normal and demented subjects. The results demonstrate high accuracy of 95%. Such a tool can be helpful in radiology image retrieval and classification.

The dramatic growth of IoT-based applications in many domains such as real-time monitoring, interactive reporting, and smart manufacturing brings challenges for adoption of cloud-based solutions for integration of latency-sensitive and resource-intensive applications. We refer to this integration as a hybrid-workflow. This paper provides a resource estimation and task scheduling framework to run hybrid workflows on edge and cloud computing systems. We propose an adaptive resource estimation technique with an online gradient descent approximation to handle the complexity of hybrid workflows. In addition, a scheduling technique to execute workflow tasks on a cooperative edge cloud system to resolve the issues of latency-sensitive application as well as to improve resource utilization at the edge layer is proposed. Experimental results show the capability of the cooperative model in reducing the time and cost of running complex and large scale hybrid workflows.

Dependence of computing resources on each other in cloud computing systems (CCS) makes them prone to fail in correlated manner which significantly impacts their service reliability and energy efficiency. Focusing on these two metrics of CCS while considering correlated failures remained an open question, which is the focus of this work. This paper proposes mechanisms for improving reliability and energy efficiency jointly under correlated failures in CCS. In order to model failure correlation, statistical cluster analysis techniques are applied to real failure traces. Then, mathematical models are built to calculate reliability and energy consumption of failure prone CCS. These mathematical models are used to design fault-tolerant and energy-aware resource provisioning mechanisms/policies. In order to further reduce the energy consumption, a correlated failure-aware VM consolidation policy is also proposed in this paper. A simulation based study of the proposed resource management policies and fault tolerance mechanisms is conducted by using real failure traces and Bag-of-Tasks workload. The results demonstrate that by exploiting failure correlation with the proposed resource management policies, we reduce the occurrence of failures on tasks by 34 percent and increase the energy efficiency of the system by 20 percent, approximately in comparison to the environments where failures are handled independently.

Smart applications, including Internet of Things (IoT) and Big Data analytics, are traditionally hosted by cloud infrastructures, which can result in high latency and cost beyond users expectation. Edge computing has emerged as a paradigm that can alleviate the pressure on clouds by delegating parts of the computation to devices in the edge of the network, at closer proximity to end users and IoT devices. In this paper, we discuss a smart application, built on top of mobile edge computing concept, to enables users to measure and analyse their food intake and support nutritional decision-making. The approach utilizes mobile edge computing to offload application computations and communications to the edge, thus saving battery life, increasing the processing capacity, and improving user comfort. In order to develop this system, we propose a loosely coupled architecture for a smart food scanner and then implement it using various IoT sensors. The performance evaluation results reveal that the implemented system can be used as an interactive appliance by users with minimum dependency and usage of their mobile phones.

The dramatic growth of the Internet of Things (IoT) technology in many application domains, ranging from intelligent video surveillance, smart retail to the Internet-of-Vehicles brings new computation challenges for rationalized utilization of computing resources. IoT application execution refers to hybrid processing model of stream and batch to achieve data analytics objectives. Hybrid workflow execution combines the challenges of latency-sensitive and resource-intensive processing. To resolve these challenges, we proposed a two stages hybrid workflow scheduling framework on edge cloud computing. In the first stage, we proposed a resource estimation algorithm based on a linear optimization approach, the gradient descent search (GDS) and in the second stage, we adopted a cluster-based provisioning and scheduling technique on heterogeneous edge cloud resources. This work provides a multi-objective optimization model for execution time and monetary cost under constraints of deadline and throughput. Results demonstrated the framework performance in controlling the execution of hybrid workflows by an efficient tuning for stream processing parameters, such as arrival rate and processing throughput. Under working constraints, the proposed scheduler provides significant improvement for large hybrid workflows in terms of execution time and monetary cost with an average of 8% and 35%, respectively.

Erasure coding is gaining attraction in cloud storage systems because it improves data reliability with huge cost savings in terms of storage. However, data recovery in erasure codes includes high disk I/O, network traffic and complex decoding that impacts degraded read latency, in case of failures. Data access latency is one of the most important metrics to determine Quality of Service. Reducing degraded latency in erasure coding is vital to improve user performance. To reduce degraded read latency of erasure codes, in this paper, we have proposed a cache based technique called ProactiveCache. This proactively copies objects in failure predicted machine into a cache tier. To deploy ProactiveCache, cloud storage system should employ various failure prediction methods to predict hardware failures. On accurate failure predictions, ProactiveCache eliminates degraded read latency. For evaluation, ProactiveCache is implemented on Ceph object storage. Experimental results show that Proactive-Cache reduces degraded read latency up to 38% and improves throughput by 37%.

The increasing adoption of Internet of Thing (IoT) technology in many application domains generates a new need for rationalized utilization of computing resources supporting such computations. IoT applications can be represented as workflows in which stream and batch applications are integrated to accomplish data analytics objectives in many application domains such as smart home, health care, bioinformatics, astronomy, education, etc. The main challenge of this combination is the differentiation of service quality constraints between the two computation paradigms. Stream processing is highly sensitive to real-time constraint while batch processes are usually resource-intensive. In this work we propose a resource provisioning framework for combined workflows which aims to find an optimal workflow configuration plan to minimize execution time and monetary cost. The framework has functions of execution plan generation, task clustering, and resource provisioning. Results show that framework is capable to control the execution of combined-workflows by efficient tunning several parameters including stream arrival rate and processing throughput.

Cloud computing provides more reliable and flexible access to IT resources, on-demand and self-service service request are some key advantages of it. Managing up-layer cloud services efficiently, while promising those advantages and SLA, motivates the challenge of provisioning and allocating resource on-demand in infrastructure layer, in response to dynamic workloads. Studies mostly have been focused on managing these demands in the physical layer and few in the application layer. This paper focuses on resource allocation method in application level that allocates an appropriate number of virtual machines to an application which requires a dynamic amount of resources. A Learning Automata based approach has been chosen to implement the method. Experimental results demonstrate that the proposed technique offers more cost effective resource provisioning approach while provisions enough resource for applications.

Virtualization as one of the leading technologies has assisted datacenters to cloudify their products and provide versatile platforms and variety of Internet services. This technology also has facilitated agile deployment of complex Internet services such as Cloud-based multi-tier applications. However implementing multi-tier applications assists providers to set up flexible and scalable services, deploying such applications in virtualized environments faces challenges, which harden collocation of disparate multi-tier applications in a shared environment. In this article, we propose a placement strategy to consolidate multi-tier applications’ tiers in a virtualized datacenter regarding workload demands of individual tiers, utilization of physical hosts’ resources and operational status of virtual machines. The placement strategy identifies over-utilized hosts based on the birth–death stochastic process and ranks tiers based on Data Envelopment Analysis-Analytic Hierarchy Process modeling to be placed on target hosts. A prioritized tier will be placed on an appropriate host though least operational interference with hosted tiers incurred. The efficiency of the proposed model is evaluated using a series of in-depth experiments. We introduce results derived from quantitative and qualitative analyses that are useful for multi-tier applications placement in virtualized datacenters. Simulation results reveal that the proposed solution excels in terms of both load distribution and energy consumption in the datacenter, while the number of unnecessary migrations and consecutive Service Level Agreement violations is considerably reduced.

Resource allocation is one of the main influential factors to provide efficient and economical processing of resources in the infrastructure as a service Clouds. While there are many challenges in providing an efficient resource allocator, maximizing the utilization of physical resources is of great importance. There are several works focused on optimizing the selection of virtual machines (VMs) for migration, however, there is less attention on the placement of the selected VMs on the available physical machines, especially for the advanced reservation request model. In this paper, the placement method and the impacts of different parameters are studied. First, different states of the problem are classified. Then, an algorithm based on integer linear programming (ILP) is proposed to solve some common cases of the problem. Finally, the algorithm is implemented in a Haizea simulator and the results are compared with the Haizea greedy algorithm and some other heuristics. The results reveal that the Haizea greedy algorithm is not able to utilize around 40% of the physical resources. Moreover, at low heterogeneity loads, the proposed ILP-based algorithm shows the same results as a Best-Fit algorithm, but at higher heterogeneity loads, the results of the proposed algorithm outperform other algorithms.

The emergence of Big Data applications provides new challenges in data management such as processing and movement of masses of data. Volunteer computing has proven itself as a distributed paradigm that can fully support Big Data generation. This paradigm uses a large number of heterogeneous and unreliable Internet-connected hosts to provide Peta-scale computing power for scientific projects. With the increase in data size and number of devices that can potentially join a volunteer computing project, the host bandwidth can become a main hindrance to the analysis of the data generated by these projects, especially if the analysis is a concurrent approach based on either in-situ or in-transit processing. In this paper, we propose a bandwidth model for volunteer computing projects based on the real trace data taken from the Docking@Home project with more than 280,000 hosts over a 5-year period. We validate the proposed statistical model using model-based and simulation-based techniques. Our modeling provides us with valuable insights on the concurrent integration of data generation with in-situ and in-transit analysis in the volunteer computing paradigm.

SUMMARY Resource provisioning is one of the main challenges in large‐scale distributed systems such as federated Grids. Recently, many resource management systems in these environments have started to use the lease abstraction and virtual machines (VMs) for resource provisioning. In the large‐scale distributed systems, resource providers serve requests from external users along with their own local users. The problem arises when there is not sufficient resources for local users, who have higher priority than external ones, and need resources urgently. This problem could be solved by preempting VM‐based leases from external users and allocating them to the local ones. However, preempting VM‐based leases entails side effects in terms of overhead time as well as increasing makespan of external requests. In this paper, we model the overhead of preempting VMs. Then, to reduce the impact of these side effects, we propose and compare several policies that determine the proper set of lease(s) for preemption. We evaluate the proposed policies through simulation as well as real experimentation in the context of InterGrid under different working conditions. Evaluation results demonstrate that the proposed preemption policies serve up to 72% more local requests without increasing the rejection ratio of external requests. Copyright © 2013 John Wiley & Sons, Ltd.

Volunteer computing systems are large-scale distributed systems with large number of heterogeneous and unreliable Internet-connected hosts. Volunteer computing resources are suitable mainly to run High-Throughput Computing (HTC) applications due to their unavailability rate and frequent churn. Although they provide Peta-scale computing power for many scientific projects across the globe, efficient usage of this platform for different types of applications still has not been investigated in depth. So, characterizing, analyzing and modeling such resources availability in volunteer computing is becoming essential and important for efficient application scheduling. In this paper, we focus on statistical modeling of volunteer resources, which exhibit non-random pattern in their availability time. The proposed models take into account the autocorrelation structure in subset of hosts whose availability has short/long-range dependence. We apply our methodology on real traces from the SETI@home project with more than 230,000 hosts. We show that Markovian arrival process can model the availability and unavailability intervals of volunteer resources with a reasonable to excellent level of accuracy.

Many applications in federated Grids have quality-of-service (QoS) constraints such as deadline. Admission control mechanisms assure QoS constraints of the applications by limiting the number of user requests accepted by a resource provider. However, in order to maximize their profit, resource owners are interested in accepting as many requests as possible. In these circumstances, the question that arises is: what is the effective number of requests that can be accepted by a resource provider in a way that the number of accepted external requests is maximized and, at the same time, QoS violations are minimized. In this paper, we answer this question in the context of a virtualized federated Grid environment, where each Grid serves requests from external users along with its local users and requests of local users have preemptive priority over external requests. We apply analytical queuing model to address this question. Additionally, we derive a preemption-aware admission control policy based on the proposed model. Simulation results under realistic working conditions indicate that the proposed policy improves the number of completed external requests (up to 25%). In terms of QoS violations, the 95% confidence interval of the average difference with other policies is between (14.79%, 18.56%).

Data-centric and service-oriented workflows are commonly used in scientific research to enable the composition and execution of complex analysis on distributed resources. Although there are a plethora of orchestration frameworks to implement workflows, most of them are not suitable to execute data-centric workflows. The main issue is transferring output of service invocations through a centralized orchestration engine to the next service in the workflow, which can be a bottleneck for the performance of a data-centric workflow. In this paper, we propose a flexible and lightweight workflow framework based on the Object Modeling Systems (OMS). Moreover, we take advantage of the OMS architecture to deploy and execute data-centric workflows in a decentralized manner to avoid passing through the centralized engine. The proposed framework is implemented in context of the Australian Urban Research Infrastructure Network (AURIN) project which is an initiative aiming to develop an e-Infrastructure supporting research in the urban and built environment research disciplines. Performance evaluation results using spatial data-centric workflows show that we can reduce 20% of the workflows execution time while using Cloud resources in the same network domain.

Resource provisiomng is an important and challenging problem in the large-scale distributed systems such as Cloud computing environments. Resource management issues such as Quality of Service (QoS) further exacerbate the resource provisioning problem. Furthermore, with the increasing functionality and complexity of Cloud computing, resource failures are inevitable. Therefore, the question we address in this paper is how to provision resources to applications in the presence of resource failures in a hybrid Cloud computing environment. To this end, we propose three Cloud resource provisioning policies where we utilize workflow applications to drive the system workload. The proposed strategies take into account the workload model and the failure correlations to redirect requests to appropriate Cloud providers. Using real failure traces and workload models, we evaluated the performance and monetary cost of the proposed policies. The results of our experiments show that we can decrease the deadline violation rate of users' requests to as low as 20% with a limited cost on Amazon public Cloud.

Multicluster systems have emerged as a promising infrastructure for provisioning of cost-effective high-performance computing and communications. Analytical models of communication networks in cluster systems have been widely reported. However, for tractability and simplicity, the existing models are based on the assumptions that the network traffic follows the nonbursty Poisson arrival process and the message destinations are uniformly distributed. Recent measurement studies have shown that the traffic generated by real-world applications reveals the bursty nature in both the spatial domain (i.e., nonuniform distribution of message destinations) and temporal domain (i.e., bursty message arrival process). In order to obtain a comprehensive understanding of the system performance, a novel analytical model is developed for communication networks in multicluster systems in the presence of the spatio-temporal bursty traffic. The spatial traffic burstiness is captured by the communication locality and the temporal traffic burstiness is modeled by the Markov-modulated Poisson process. After validating its accuracy through extensive simulation experiments, the model is used to investigate the impact of bursty message arrivals and communication locality on network performance. The analytical results demonstrate that the communication locality can relieve the degrading effects of bursty message arrivals on the network performance.

International audience

With the increasing functionality and complexity of distributed systems, resource failures are inevitable. While numerous models and algorithms for dealing with failures exist, the lack of public trace data sets and tools has prevented meaningful comparisons. To facilitate the design, validation, and comparison of fault-tolerant models and algorithms, we have created the Failure Trace Archive (FTA) as an online public repository of availability traces taken from diverse parallel and distributed systems. Our main contributions in this study are the following. First, we describe the design of the archive, in particular the rationale of the standard FTA format, and the design of a toolbox that facilitates automated analysis of trace data sets. Second, applying the toolbox, we present a uniform comparative analysis with statistics and models of failures in nine distributed systems. Third, we show how different interpretations of these data sets can result in different conclusions. This emphasizes the critical need for the public availability of trace data and methods for their analysis.

Performance evaluation is a serious challenge in designing or optimizing reconfigurable instruction set processors. The conventional approaches based on synthesis and simulations are very time consuming and need a considerable design effort. A combined analytical and simulation-based model (CAnSO*) is proposed and validated for performance evaluation of a typical reconfigurable instruction set processor. The proposed model consists of an analytical core that incorporates statistics gathered from cycle-accurate simulation to make a reasonable evaluation and provide a valuable insight. Compared to cycle-accurate simulation results, CAnSO proves almost 2% variation in the speedup measurement.

Cluster-based systems have emerged as a promising technology for providing cost-effectiveness in high-performance computing and communication systems. Performance studies on communication networks in cluster-based systems have been reported based on the simplified assumptions that the traffic follows the non-bursty Poisson process and the message destinations are uniformly distributed over all network nodes. However, the uniform distribution of message destinations is not always realistic in practice. Moreover, the communication locality, a typical example of the non-uniform destination distribution, has been shown to be an important phenomenon in the communication networks of cluster systems. Many recent measurement studies have revealed that the traffic generated by many real-world applications exhibits a high degree of burstiness. In order to have a comprehensive understanding of the system performance, this paper proposes a new analytical model for communication networks in multi-cluster systems under the bursty message arrivals with communication locality. The model is validated through extensive simulation experiments.

Cloud Computing has taken commercial computing by storm. However, adoption of cloud computing platforms and services by the scientific community is in its infancy as the performance and monetary cost-benefits for scientific applications are not perfectly clear. This is especially true for desktop grids (aka volunteer computing) applications. We compare and contrast the performance and monetary cost-benefits of clouds for desktop grid applications, ranging in computational size and storage. We address the following questions: (i) What are the performance tradeoffs in using one platform over the other? (ii) What are the specific resource requirements and monetary costs of creating and deploying applications on each platform? (iii) In light of those monetary and performance cost-benefits, how do these platforms compare? (iv) Can cloud computing platforms be used in combination with desktop grids to improve cost-effectiveness even further? We examine those questions using performance measurements and monetary expenses of real desktop grids and the Amazon elastic compute cloud.

In the age of cloud, Grid, P2P, and volunteer distributed computing, large-scale systems with tens of thousands of unreliable hosts are increasingly common. Invariably, these systems are composed of heterogeneous hosts whose individual availability often exhibit different statistical properties (for example stationary versus non-stationary behavior) and fit different models (for example Exponential, Weibull, or Pareto probability distributions). In this paper, we describe an effective method for discovering subsets of hosts whose availability have similar statistical properties and can be modelled with similar probability distributions. We apply this method with about 230,000 host availability traces obtained from a real large-scale Internet-distributed system, namely SETI@home. We find that about 34% of hosts exhibit availability that is a truly random process, and that these hosts can often be modelled accurately with a few distinct distributions from different families. We believe that this characterization is fundamental in the design of stochastic scheduling algorithms across large-scale systems where host availability is uncertain.

Abstract The trends in parallel processing system design and deployment have been toward networked distributed systems such as cluster computing systems. Since the overall performance of such distributed systems often depends on the efficiency of their communication networks, performance analysis of the interconnection networks for such distributed systems is paramount. In this paper, we develop an analytical model, under non‐uniform traffic and in the presence of communication locality, for the m‐port n‐tree family interconnection networks commonly employed in large‐scale cluster computing systems. We use the proposed model to study two widely used interconnection networks flow control mechanism namely the wormhole and store&forward. The proposed analytical model is validated through comprehensive simulation. The results of the simulation demonstrated that the proposed model exhibits a good degree of accuracy for various system organizations and under different working conditions. Copyright © 2007 John Wiley & Sons, Ltd.

The study of the communication networks for distributed systems is very important, since the overall performance of these systems is often depends on the effectiveness of its communication network. In this paper, we address the problem of networks modeling for heterogeneous large-scale cluster systems. We consider the large-scale cluster systems as a typical cluster of clusters system. Since the heterogeneity is becoming common in such systems, we take into account network as well as cluster size heterogeneity to propose the model. To this end, we present an analytical network model and validate the model through comprehensive simulation. The results of the simulation demonstrated that the proposed model exhibits a good degree of accuracy for various system organizations and under different working conditions

Execution of MPI applications on clusters and grid deployments suffers from node and network failure that motivates the use of fault tolerant MPI implementations. Two category techniques have been introduced to make these systems fault-tolerant. The first one is checkpoint-based technique and the other one is called log-based recovery protocol. Sender-based pessimistic logging which falls in the second category is harnessing from huge amount of messages payloads which must be kept in volatile memory. In this paper, we present a coordinated checkpoint from message payload (CCMP) to reduce the aforementioned overhead. The proposed method was examined by MPICH-V2, a public domain platform implementing pessimistic logging with uncoordinated checkpoint. Experimental results demonstrated the reduction of run-time for NPB benchmarks in both fault-free and faulty environments.

This paper presents an area efficient, low-power and robust ACS unit for Viterbi decoder in two synchronous and asynchronous architectures. The asynchronous design is based upon quasi delay insensitive (QDI) timing model which leads to a robust and low power purpose and synchronous architecture uses a hybrid CMOS-pseudo NMOS technology to improve area and throughput factors. Some optimization techniques to reduce the power and area are applied to each design. The simulation results show the asynchronous design has the lowest power consumption with 6.65mW and hybrid CMOS has the lowest transistor counts with 759 in relative to other reported circuits.