Reliable localization under Global Positioning System-denied or visually degraded conditions remains a fundamental challenge for autonomous systems. Vision- and Light Detection and Ranging (LiDAR)-based approaches often degrade in low illumination, adverse weather, or appearance-changing environments, as they rely on stable surface texture or geometry. In contrast, ground-penetrating radar (GPR) captures subsurface electromagnetic reflections that remain relatively stable across lighting, seasonal, and weather variations, making it a promising complementary sensing modality for long-term localization. However, spatial variability in subsurface dielectric properties induces fluctuations in electromagnetic wave velocity, leading to geometric distortions in GPR echoes and unstable feature extraction. To address this challenge, we propose the Velocity-Invariant Feature Transform (VIFT), a physics-guided self-supervised learning framework for GPR keypoint detection and description. VIFT explicitly models wave-velocity-induced distortions through a continuous velocity space parameterized by a Beta distribution, and leverages velocity-conditioned wavefield migration as physically consistent data augmentation. A Siamese network is trained with velocity-consistency supervision to jointly learn repeatable keypoint score maps and discriminative local descriptors from unlabeled real GPR scans. To further enhance robustness, sparsity-aware, dispersion, distinctiveness, and orthogonality losses are incorporated to improve repeatability, spatial coverage, and descriptor discriminability. Extensive experiments on public benchmarks and large-scale real-world GPR datasets demonstrate that VIFT consistently outperforms traditional handcrafted methods and recent learning-based Vison and GPR methods, achieving a 5–10% improvement in keypoint repeatability over state-of-the-art methods, particularly under extremely sparse keypoint sampling regimes, while also improving matching accuracy and registration robustness under diverse subsurface conditions.

Current methods for 3D semantic segmentation propose training models with limited annotations to address the difficulty of annotating large, irregular, and unordered 3D point cloud data. They usually focus on the 3D domain only, without leveraging the complementary nature of 2D and 3D data. Besides, some methods extend original labels or generate pseudo labels to guide the training, but they often fail to fully use these labels or address the noise within them. Meanwhile, the emergence of comprehensive and adaptable foundation models has offered effective solutions for segmenting 2D data. Leveraging this advancement, we present a novel approach that maximizes the utility of sparsely available 3D annotations by incorporating segmentation masks generated by 2D foundation models. We further propagate the 2D segmentation masks into the 3D space by establishing geometric correspondences between 3D scenes and 2D views. We extend the highly sparse annotations to encompass the areas delineated by 3D masks, thereby substantially augmenting the pool of available labels. Furthermore, we apply confidence- and uncertainty-based consistency regularization on augmentations of the 3D point cloud and select the reliable pseudo labels, which are further spread on the 3D masks to generate more labels. This innovative strategy bridges the gap between limited 3D annotations and the powerful capabilities of 2D foundation models, ultimately improving the performance of 3D weakly supervised segmentation.

Depth cues are essential for visual perception tasks like Salient Object Detection (SOD). Due to varying depth reliability across scenes, some researchers propose evaluating the overall quality of the depth maps and discarding the less reliable ones to avoid contamination. However, these methods often fail to fully utilize valuable information in depth maps, leading to sub-optimal performance particularly when depth quality is unreliable. Since low-quality depth maps still contain useful information that potentially improves model performance, we propose a Depth Pixel-wise Potential-aware Network to leverage these depth cues effectively. This network includes two novel components designed: 1) A learning strategy for explicitly modeling the confidence of each depth pixel to assist the model in locating valid information in the depth map. 2) A cross-modal adaptive multiple fusion module that fuses features from both RGB and depth modalities. It aims to mitigate the contamination effect of unreliable depth maps and fully exploit the benefits of multiple fusion strategies. Experimental results show that on four publicly available datasets, our method outperforms 17 mainstream methods on various evaluation metrics.

Domain adaptive object detection aims to adapt detection models to domains where annotated data is unavailable. Existing methods have been proposed to address the domain gap using the semi-supervised student-teacher framework. However, a fundamental issue arises from the class imbalance in the labelled training set, which can result in inaccurate pseudo-labels. The relationship between classes, especially where one class is a majority and the other minority, has a large impact on class bias. We propose Class-Aware Teacher (CAT) to address the class bias issue in the domain adaptation setting. In our work, we ap-proximate the class relationships with our Inter-Class Relation module (ICRm) and exploit it to reduce the bias within the model. In this way, we are able to apply augmentations to highly related classes, both inter- and intra-domain, to boost the performance of minority classes while having minimal impact on majority classes. We further reduce the bias by implementing a class-relation weight to our classification loss. Experiments conducted on various datasets and ablation studies show that our method is able to address the class bias in the domain adaptation setting. On the Cityscapes$\rightarrow$Foggy Cityscapes dataset, we attained a 52.5 mAp, a substantial improvement over the 51.2 mAP achieved by the state-of-the-art method.11www.github.com/mecarill/cat

Self-supervised frameworks for representation learning have recently stirred up interest among the remote sensing community, given their potential to mitigate the high labeling costs associated with curating large satellite image datasets. In the realm of multimodal data fusion, while contrastive learning methods can help bridge the domain gap between different sensor types, they rely on data augmentation techniques that require expertise and careful design, especially for multispectral remote sensing data. A possible but rather scarcely studied way to circumvent these limitations is to use a masked image modelling based pretraining strategy. In this paper, we introduce Fus-MAE, a self-supervised learning framework based on masked autoencoders that uses cross-attention to perform early and feature-level data fusion between synthetic aperture radar and multispectral optical data - two modalities with a significant domain gap. Our empirical findings demonstrate that Fus-MAE can effectively compete with contrastive learning strategies tailored for SAR-optical data fusion and outperforms other masked-autoencoders frameworks trained on a larger corpus. For replicability, code and weights are provided in this github repository.

Image Alignment between Synthetic Aperture Radar (SAR) and Electro-Optical (EO) imagery is a task that has comprehensive remote sensing capabilities. Traditional deep learning-based SAR-optical image matching models heavily rely on supervised learning with expensive annotated datasets, leading to reduced accuracy and overfitting when encountering insufficient data during training. To tackle this issue, this paper proposes a student-teacher framework for Domain Adaptation (DA) approach, transferring deep learning models from well-annotated source domains like normal outside optical images to non-annotated SAR-EO target domains. Additionally, in contrast to previous methods which usually use CNN or ordinary Transformer structure to extract features from image pairs, we use self and cross attention mechanisms in Transformer to obtain feature descriptors that are conditioned on both multimodal images. The larger global receptive field and better feature extraction capability provided by this Transformer shows its ability to accommodate large disparities of multimodal data and manage fewer textures such as rural areas with forest or desert in satellite images, where previous backbones usually struggle to produce repeatable and correct interest points.

Object detection at night is a challenging problem due to the absence of night image annotations. Despite several domain adaptation methods, achieving high-precision results remains an issue. False-positive error propagation is still observed in methods using the well-established student-teacher framework, particularly for small-scale and low-light objects. This paper proposes a two-phase consistency unsupervised domain adaptation network, 2PCNet, to address these issues. The network employs high-confidence bounding-box predictions from the teacher in the first phase and appends them to the student's region proposals for the teacher to re-evaluate in the second phase, resulting in a combination of high and low confidence pseudolabels. The night images and pseudo-labels are scaled-down before being used as input to the student, providing stronger small-scale pseudo-labels. To address errors that arise from low-light regions and other night-related attributes in images, we propose a night-specific augmentation pipeline called NightAug. This pipeline involves applying random augmentations, such as glare, blur, and noise, to daytime images. Experiments on publicly available datasets demonstrate that our method achieves superior results to state-of-the-art methods by 20%, and to supervised models trained directly on the target data.11www.github.com/mecarill/2pcnet

The use of deep learning technology to manipulate images and videos of people in ways that are difficult to distinguish from the real ones, known as deepfake, has become a matter of national security concern in recent years. As a result, many studies have been carried out to detect deepfake and manipulated media. Among these studies, deep video models based on convolutional neural networks have been the preferred method for detecting deepfake in videos. This study presents a novel deep video model called Sequential-Parallel Networks (SPNet) that provides efficient deepfake detection. The SPNet model consists of a simple yet innovative sequential-parallel block that first extracts spatial and temporal features sequentially, then concatenates them together in parallel. As a result, the presented SPNet possesses comparable spatiotemporal modeling abilities as most state-of-the-art deep video methods but with lower computation complexity and fewer parameters. The efficiency of the presented SPNet is demonstrated on a large-scale deepfake benchmark in terms of high recognition accuracy and low computational cost.

Supervised learning methods have demonstrated state-of-the-art performance for spiking neural network (SNN) but require a large amount of annotated training data, which may be expensive to obtain. To address this, we propose a semi-supervised learning method based on pseudo-labeling that enables SNN training using a small number of annotated training samples. Our proposed method outperforms the existing semi-supervised learning approaches for SNN by more than 17% on the MNIST dataset when 100 annotated samples are used. Our proposed spike-based method is hardware friendly, can be incorporated to various SNN models, and does not involve intensive computations, which is desirable for applications on internet of things (IoT) devices that have tight constraints on hardware resources and energy consumption.

The applications that are deployed in the cloud to provide services to the users encompass a large number of interconnected dependent cloud components. Multiple identical components are scheduled to run concurrently in order to handle unexpected failures and provide uninterrupted service to the end user, which introduces resource overhead problem for the cloud service provider. Furthermore such resource-intensive fault tolerant strategies bring extra monetary overhead to the cloud service provider and eventually to the cloud users. In order to address these issues, a novel fault tolerant strategy based on the significance level of each component is developed. The communication topology among the application components, their historical performance, failure rate, failure impact on other components, dependencies among them, etc., are used to rank those application components to further decide on the importance of one component over others. Based on the rank, a Markov Decision Process (MDP) model is presented to determine the number of replicas that varies from one component to another. A rigorous performance evaluation is carried out using some of the most common practically useful metrics such as, recovery time upon a fault, average number of components needed, number of parallel components successfully executed, etc., to quote a few, with similar component ranking and fault tolerant strategies. Simulation results demonstrate that the proposed algorithm reduces the required number of virtual and physical machines by approximately 10% and 4.2%, respectively, compared to other similar algorithms.

Nowadays, increasing number of services are provided to individuals and organizations through cloud computing systems in apay-as-you-usemodel. This business service paradigm encounters several cloud Quality of Service (QoS) challenges, such as reliability, cost, and response time. The most common mechanism to improve cloud service reliability is a primary/backup (PB) fault-tolerant technique. However, this reliability enhancement technique inevitably results in multiple replications, which lead to high service cost. In recognition of these challenges, we first build a cloud computing systems resources management architecture. Then, we analyze the cloud service execution reliability on the physical resources of a VM and used a CUDA (Compute Unified Device Architecture)-enabled parallel two-dimensional long short-term memory neural network to predict the software faults of a cloud VM. Third, we propose an effective primary/backup cloud service cost calculation approach. To overcome the cloud service response time constraint, we integrate a response time slack factor into this method. Fourth, we formulate the cloud service reliability and cost aware job scheduling problem, which aims at minimizing the total cloud service cost and rejection rate, and improving the system reliability. Fifthly, a heuristic greedy reliability and cost aware job scheduling (RCJS) algorithm is proposed. Finally, a performance evaluation is conducted and the experimental results demonstrate that our proposed RCJS algorithm significantly outperforms optimal redundant VM placement (OPVMP), MIN-MIN algorithms in terms of average service cost and rejection rate. This algorithm also demonstrates good trade-off of reliability when compared to the other two algorithms and is suitable for cloud services with high reliability and low-cost requirements.

Microservices are an architectural and organizational paradigm for Internet application development. In cloud data centers, delay-sensitive applications receive massive user requests, which are fed into multiple queues and subsequently served by multiple microservice instances. Accordingly, effective deployment of multiple queues and containers can significantly reduce queuing delay, processing delay, and communication delay. Due to the increased complexity of call dependencies and probabilistic routing paths, the deployment of service instances fully interacts with request routing, bringing great difficulties to service orchestration. In this case, it is valuable to simultaneously consider service deployment and request routing in a fine-grained manner. However, most existing studies considered them as two independent components with local optimization, while data dependencies and the instance-level deployment are ignored. Therefore, this paper proposes to jointly optimize the deployment and request routing of microservice call graphs based on fine-grained queuing network analysis and container orchestration. We first formulate the problem as a mixed-integer nonlinear program and exploit open Jackson queuing networks to model intrinsic data dependencies and analyze response latency. To optimize the overall cost and latency, this paper presents an efficient two-stage heuristic algorithm, which consists of a resource-splitting-based deployment approach and a partition-mapping-based routing method. Further, this paper also provides mathematical analysis on the performance and complexity of the proposed algorithm. Finally, comprehensive trace-driven experiments demonstrate that the overall performance of our approach is better than existing microservice benchmarks. The average deployment cost is reduced by 27.4% and end-to-end response latency is reduced by 15.1% on average.

Fog Computing usefully extends Cloud to the edge of the network for the sake of meeting users’ expanding demand for low latency. However, due to its scattered distribution and open architecture, fog nodes are highly vulnerable to security threats, resulting in an inevitable sharp conflict between quick response time and high data security. This conflict motivates the need for effective data placement among fog nodes towards a trade-off between security and time. Existing studies merely offer independent solutions by considering either security or response time. By contrast, we establish a dynamic multi-objective optimization model in this article by optimizing security and response time simultaneously. With this model, we propose an efficient evolutionary algorithm, referred to asDynamic Interactive Security-and-Time cognizant algorithm(DIST), to obtain optimal data placement strategies under Fog environments. To improve efficiency,DISTallows users to gradually incorporate their preference information into the search process so as to find their most preferred solutions without exploring the whole search space. We demonstrate the superiority ofDISTby rigorous comparison with the most state-of-art data placement strategy and other well-applied strategies. Experimental results manifest thatDISToutperforms other strategies in obtaining solutions with higher data security and shorter response time. Furthermore,DISTis capable of efficiently and continuously tracking the Pareto optimal solution under dynamically changing Fog environments while other existing strategies cannot.

Multi-label Learning on image data has been widely exploited with deep learning models. However, supervised training on deep CNN models often cannot discover sufficient discriminative features for classification. As a result, numerous self-supervision methods are proposed to learn more robust image representations. However, most self-supervised approaches focus on single-instance single-label data and fall short on more complex images with multiple objects. Therefore, we propose an Object-Aware Self-Supervision (OASS) method to obtain more fine-grained representations for multi-label learning, dynamically generating auxiliary tasks based on object locations. Secondly, the robust representation learned by OASS can be leveraged to efficiently generate Class-Specific Instances (CSI) in a proposal-free fashion to better guide multi-label supervision signal transfer to instances. Extensive experiments on the VOC2012 dataset for multi-label classification demonstrate the effectiveness of the proposed method against the state-of-the-art counterparts.

With large-scale well-labeled datasets, deep learning has shown significant success in medical image segmentation. However, it is challenging to acquire abundant annotations in clinical practice due to extensive expertise requirements and costly labeling efforts. Recently, contrastive learning has shown a strong capacity for visual representation learning on unlabeled data, achieving impressive performance rivaling supervised learning in many domains. In this work, we propose a novel multi-scale multi-view global-local contrastive learning (MMGL) framework to thoroughly explore global and local features from different scales and views for robust contrastive learning performance, thereby improving segmentation performance with limited annotations. Extensive experiments on the MM-WHS dataset demonstrate the effectiveness of MMGL framework on semi-supervised cardiac image segmentation, outperforming the state-of-the-art contrastive learning methods by a large margin.

While deep models have shown promising performance in medical image segmentation, they heavily rely on a large amount of well-annotated data, which is difficult to access, especially in clinical practice. On the other hand, high-accuracy deep models usually come in large model sizes, limiting their employment in real scenarios. In this work, we propose a novel asymmetric co-teacher framework, ACT-Net, to alleviate the burden on both expensive annotations and computational costs for semi-supervised knowledge distillation. We advance teacher-student learning with a co-teacher network to facilitate asymmetric knowledge distillation from large models to small ones by alternating student and teacher roles, obtaining tiny but accurate models for clinical employment. To verify the effectiveness of our ACT-Net, we employ the ACDC dataset for cardiac substructure segmentation in our experiments. Extensive experimental results demonstrate that ACT-Net outperforms other knowledge distillation methods and achieves lossless segmentation performance with 250× fewer parameters.

In this paper, we report on the design of a neuromorphic co-processor on a Field Programmable Gate Array (FPGA) platform that is capable of emulating Spiking Neural Networks (SNNs) with support for on-chip unsupervised learning. One defining feature of our design is that the SNN configuration is defined entirely in the software executed by our neuromorphic co-processor. Evaluation on the FPGA platform shows that our design consumes a small amount of hardware resources and on-chip memory storage (438.75 kB). In addition, the inference and the on-chip learning in a deep convolutional SNN emulated on our FPGA implementation are $10.5vf \times$ and $8.6 \times$ faster than the implementation on high performance x86 CPU. Moreover, we demonstrate the ability of our neuromorphic co-processor to perform the on-chip learning on an object recognition task (based on the Caltech-101 dataset).

Processing a large-scale Synthetic Aperture Radar (SAR) image dataset on a distributed computing infrastructure poses a challenging problem. Large-scale load distribution strategies like multi-installment scheduling (MIS) assume that the size of the result is negligible compared to the input workloads and hence ignore it in their design. Similarly, numerical methods like particle swarm optimization and their variants are not practical for real-time applications, given their run-time complexities. As both the results retrieval and completion time are crucial for SAR image data processing, in this article, we attempt to provide a thorough theoretical analysis of an adaptive MIS that includes the result retrieval phase. We use the periodic nature of the internal installments to keep the strategy simple and fine-tune the last installment to avoid any idle times in the processors. We derive a closed-form solution for the load fractions and hence, the overall processing time, schedule feasibility criteria, and certain other properties that lead to adaptive scheduling. Finally, we validate our theoretical findings through rigorous simulation studies using a loosely connected virtual machines (VMs) topology for the SAR dataset.

When a set of self-interested users shares multiple resources in a distributed system, we face the problem of allocating resources, called the load balancing problem. In particular, load balancing is defined as allocating the load to the servers of the distributed system such that jobs’ response time is minimized, and the utilization of servers is improved. In this article, the load balancing problem in a distributed system consists of a finite set of servers, and a finite set of users is studied. The load balancing problem considered here is a bi-objective problem with two highly probable conflicting objectives: (i) minimizing jobs’ response time (ii) providing the fair utilization of servers. In order to satisfy these two objectives simultaneously, both the objectives are considered in an integrated manner. Next, the load balancing problem is formulated as a noncooperative game; and to solve the game (i.e., to find the Nash equilibrium), a distributed load balancing algorithm (DLBA) is proposed. An experimental study is carried out to ascertain the efficacy of the proposed DLBA. Further, we compare DBLA with three existing load balancing approaches to evaluate its comparative effectiveness. The experimental results validate the effectiveness of the DLBA over the existing approaches.

Application-level approximate computing exploits inherent resilience of adaptive applications, and trades off application output quality for runtime system resources. Existing methods treat computing quality as the number of clock cycles to execute a task, but they overlook the fact that the quality of many real-life applications exhibit the characteristic of diminishing return as the processor continues executing. The diminishing return of the quality is largely due to the features of iterative processing or successive refinement inherent in those applications. Ignoring it leads to large over-estimation in contemporary quality optimization approaches. In this article, we exploit the application adaptability to achieve quality maximization by taking both system resource constraints and diminishing return of the quality into account. We first reveal that the diminishing return of the quality is inherent in several well-known applications, and suggest an exponential model that accurately captures it. Second, we propose a dynamic frequency scaling (DFS) methodology to optimally decide the processor execution cycles for such applications, in order to maximize the output quality under system energy, timing, and temperature constraints. We transform the DFS problem to an iterative pseudo quadratic programming heuristic that can be efficiently solved. Third, we present a wrapping dynamic voltage scaling (wDVS) methodology to achieve further quality improvement, by judiciously adjusting the supply voltage to provide extra frequency scaling space. Compared to state-of-the-art algorithms, our approach produces at least 19.1 percent quality improvement on all evaluated cases, with negligible execution overhead.

Fully homomorphic encryption (FHE) is a powerful cryptographic primitive to secure outsourced computations against an untrusted third-party provider. With the growing demand for AI and the usefulness of machine learning as a service (MLaaS), the need for secure training and inference of artificial neural networks is rising. However, the computational complexity of existing FHE schemes has been a strong deterrent to this. Prior works suffered from accuracy degradation, lack of scalability, and ciphertext expansion issues. In this paper, we take the first step towards the problem of space-efficiency in evaluating deep neural networks through designing DOReN: a low depth, batched neuron that can simultaneously evaluate multiple quantized ReLU-activated neurons on encrypted data without approximations. Our circuit design reduced the complexity of the accumulator circuit depth from O(logm ·logn) to O(logm + logn) for n bit integers. The experimental results show that the amortized processing time of our homomorphic neuron is approximately 1.26 seconds for 300 inputs and less than 0.13 seconds for 10 inputs at 80 bit security, which is a 20 fold improvement upon Lou and Jiang, NeurIPS 2019.

We present a multi-GPU design, implementation and performance evaluation of the Halevi-Polyakov-Shoup (HPS) variant of the Fan-Vercauteren (FV) levelled Fully Homomorphic Encryption (FHE) scheme. Our design follows a data parallelism approach and uses partitioning methods to distribute the workload in FV primitives evenly across available GPUs. The design is put to address space and runtime requirements of FHE computations. It is also suitable for distributed-memory architectures, and includes efficient GPU-to-GPU data exchange protocols. Moreover, it is user-friendly as user intervention is not required for task decomposition, scheduling or load balancing. We implement and evaluate the performance of our design on two homogeneous and heterogeneous NVIDIA GPU clusters: K80, and a customized P100. We also provide a comparison with a recent shared-memory-based multi-core CPU implementation using two homomorphic circuits as workloads: vector addition and multiplication. Moreover, we use our multi-GPU Levelled-FHE to implement the inference circuit of two Convolutional Neural Networks (CNNs) to perform homomorphically image classification on encrypted images from the MNIST and CIFAR - 10 datasets. Our implementation provides 1 to 3 orders of magnitude speedup compared with the CPU implementation on vector operations. In terms of scalability, our design shows reasonable scalability curves when the GPUs are fully connected.

The Hadoop enabled cloud platforms are gradually becoming preferred computational environment to execute scientific big data workloads in a periodic manner. However, it is observed that the default data placement approach of such cloud platforms is not the efficient one and often ends up with significant data transfer overhead leading to degradation of the overall job completion time. In this article, a Resource and Network-aware Data Placement Algorithm (RENDA) is proposed to reduce the non-local executions and thereby reduce the overall job completion time for periodic workloads in the cloud environment. The entire job execution is modeled as a two-stage execution characterized as data distribution and data processing. The RENDA reduces the time of the stages as mentioned above by estimating the heterogeneous performance of the nodes on a real-time basis followed by careful allocation of data in several installments to participating nodes. The experimental results show that the proposed RENDA algorithm consistently outperforms over the recent state-of-the-art alternatives with as much as 28 percent reduction in data transfer overhead leading to 16 percent reduction in average job completion time with 27 percent average speedup on average job execution.

Cross-layer reliability (CLR) presents a cost-effective alternative to traditional single-layer design in resource-constrained embedded systems. CLR provides the scope for leveraging the inherent fault-masking of multiple layers and exploiting application-specific tolerances to degradation in some Quality of Service (QoS) metrics. However, it can also lead to an explosion in the design complexity. State-of-the art approaches to such joint optimization across multiple degrees of freedom can lead to degradation in the system-level Design Space Exploration (DSE) results. To this end, we propose a DSE methodology for enabling CLR-aware task-mapping in heterogeneous embedded systems. Specifically, we present novel approaches to both task and system-level analysis for performing an early-stage exploration of various design decisions. The proposed methodology results in considerable improvements over other state-of-the-art approaches and shows significant scaling with application size.

Nowadays, cloud-supported Internet of Things (Cloud-IoT) has been broadly deployed in smart medical systems, where the limitations of Internet of Things (IoT)-associated medical devices in terms of data access, storage, scalability, and computing are solved through the use of cloud computing architectures. However, with the rapid development of medical equipment and the increasing number of medical devices, it will be extremely difficult to program or manage such an expanding and massive medical IoT system in traditional single-cloud platforms. In this article, we design and implement a multicloud framework for building OpenStack-based platform for medical IoT, referred to as the tri-storage failure recovery system (Tri-SFRS). To implement Tri-SFRS, we combine several techniques to achieve this reduction in effort, including a multicloud cascading architecture, a low-overhead native testing framework, a medical data storage-backup mechanism, and snapshot-volume cascaded operations for b-ultrasonic data. Tri-SFRS is also able to simultaneously enable resource management specialization. Tri-SFRS has been designed as a native component in the OpenStack platform, and it demonstrates the degree of native OpenStack multicloud platform management by our proposed cascading framework. Comparing with the traditional single-cloud OpenStack platform, Tri-SFRS can reduce the resource-request processing latency from B ultrasonic machines by up to 20%. Our experiments also demonstrate the broad applicability of Tri-SFRS.

Technology scaling and architectural innovations have led to increasing ubiquity of embedded systems across applications with widely varying and often constantly changing performance and reliability specifications. However, the increasing physical fault-rates in electronic systems have led to single-layer reliability approaches becoming infeasible for resource-constrained systems. Dynamic Cross-layer reliability (CLR) provides scope for efficient adaptation to such QoS variations and increasing unreliability. We propose a design methodology for enabling QoS-aware CLR-integrated runtime adaptation in heterogeneous MPSoC-based embedded systems. Specifically, we propose a combination of reconfiguration cost-aware optimization at design-time and an agent-based optimization at run-time. We report a reduction of up to 51% and 37% in average reconfiguration cost and average energy consumption respectively over state-of-the-art approaches.

Multi-installment scheduling (MIS) has been deemed as a promising paradigm that can sharply reduce the processing time of large-scale divisible workloads on various network-based compute platforms. Unfortunately, the practicality of MIS was crippled due to its overwhelming complexity for deriving optimal values for (n × m) + 2 related variables, i.e., we have to obtain an optimal number n of required computing resources, optimal number m of installments, and optimal load partition matrix A = (αij)n×m which determines the sizes of load fractions assigned to each computing unit in every installment. To circumvent this complexity, in this paper, we first derive explicit analytical expressions for optimal load partition matrix A of size n × m based on a given number of n and m. Then we propose a heuristic algorithm referred to as Time, Resource, and Energy Efficient MIS (TREE-MIS) to determine optimal values of n and m. The efficiency of our approach is shown to significantly improve since it can produce globally optimal solutions directly for (n × m) variables among (n × m) + 2 in total for MIS problems based on the derived analytical expressions within a short runtime. We conduct extensive simulations to demonstrate the effectiveness of the proposed algorithm. Simulation results show that our TREE-MIS can not only minimize the processing time of workloads as well as improve resource utilization of the compute platform but also drastically reduce the runtime compared to other state-of-art MIS strategies. Furthermore, while handling large-scale workloads in any large network infrastructures would inexorably result in significant amounts of energy wastage if the strategy is not prudently designed. As an offshoot of our analysis and design, we clearly demonstrate that the energy wastage in adopting our TREE-MIS is kept minimum when compared to other currently available strategies in practice.

Dynamic Partial Reconfiguration (DPR) in reconfigurable platforms can be used for the mitigation of aging-related permanent faults. We propose an application-specific system-level design methodology for determining the appropriate number of Partially Reconfigurable Regions and their compatibility with Partially Reconfigurable Modules for maximizing the system lifetime. Specifically, we propose a lifetime-aware scheduler that maximizes system MTTF. We use the scheduler along with an automated floorplanner for design space exploration at design-time to generate a heterogeneous PRR system. Our experiments show that the heterogeneous systems can offer up to 2x lifetime improvement over homogeneous ones.

Dynamic Partial Reconfiguration (DPR) can be used for time-sharing of computing resources within Partially Reconfigurable Regions (PRRs) in FPGA-based systems. The heterogeneous partitioning in such systems allows the user to exploit the application-specific mapping of Partially Reconfigurable Modules (PRMs) to PRRs to implement more efficient designs. It offers increased opportunities in optimizing the reliability of the system across multiple layers - from the low-level physical one to the higher application layer. This method, called cross-layer reliability, can potentially exploit the application-specific tolerances to the quality of service (QoS) to tackle the increasing device fault-rates more cost-effectively by distributing the fault-mitigation to different layers. In this work, we propose a QoS-aware cross-layer reliability-integrated design methodology for FPGA-based DPR systems. Specifically, our methodology analyzes the requirements of the applications in terms of Functional Reliability, System Lifetime and Makespan to determine the best possible combinations of reliability-oriented design choices in different layers. We report up to an average of 24% and 30% performance improvements for single and multi-objective optimization-based system partitioning.

Outsourcing data to a third-party administrative control, as is done in cloud computing, gives rise to security concerns. The data compromise may occur due to attacks by other users and nodes within the cloud. Therefore, high security measures are required to protect data within the cloud. However, the employed security strategy must also take into account the optimization of the data retrieval time. In this paper, we propose division and replication of data in the cloud for optimal performance and security (DROPS) that collectively approaches the security and performance issues. In the DROPS methodology, we divide a file into fragments, and replicate the fragmented data over the cloud nodes. Each of the nodes stores only a single fragment of a particular data file that ensures that even in case of a successful attack, no meaningful information is revealed to the attacker. Moreover, the nodes storing the fragments, are separated with certain distance by means of graph T-coloring to prohibit an attacker of guessing the locations of the fragments. Furthermore, the DROPS methodology does not rely on the traditional cryptographic techniques for the data security; thereby relieving the system of computationally expensive methodologies. We show that the probability to locate and compromise all of the nodes storing the fragments of a single file is extremely low. We also compare the performance of the DROPS methodology with 10 other schemes. The higher level of security with slight performance overhead was observed.

Virtualization technology boosts up traditional computing concept to cloud computing by introducing Virtual Machines (VMs) over the Physical Machines (PMs), which enables the cloud service providers to share the limited computing and network resources among multiple users. Virtual resource mapping can be defined as the process of embedding multiple VMs and their network resource demand onto multiple inter-connected PMs. The existing mechanisms of resource mapping need to be efficient enough to minimize the number of PMs without compromising the deadline of the tasks assigned to the VMs, which is NP-hard. To deal with this problem, a Link based Virtual Resource Management (LVRM) algorithm is designed to map the VMs onto PMs based on the available and required resources of the PMs and VMs, respectively. The designed algorithm exploits the fact that the demanded network bandwidth among VMs should be given higher priority while allocating the physical resources to the inter-connected virtual machines as insufficient network bandwidth may detain the task execution. The proposed algorithm is evaluated by a discrete event simulator and is compared with similar virtual network embedded algorithms. Simulation results show that LVRM can outperform over other network embedded algorithms.

Nowadays, more and more online services based on cloud computing have taken the places of some traditional applications (e.g., Health Care) that continuously generate large volume of data and require data storage and analysis in time. Such applications can be categorized as “Continuous Writing Applications” (CWA) that have particular requirements on bandwidth, storage, computation, and service reliability. In the meanwhile, they are very sensitive to the cost. In this paper, we present an architecture of multiple cloud service providers (CSPs) or “Cloud-of-Clouds” to provide services to the CWA and propose a novel resource scheduling algorithm to minimize the cost of entire systems. Difference from many research efforts that focus on a single resource, we take many factors into considerations that include user's requirements of bandwidth, storage and computation, the resources of CSPs that can provide, CSPs for data backup, the configurations of Cloud-of-Clouds, system models of CSPs, and many more. We first present the system models of classic CWA applications to capture the resource requirements of users on Cloud-of-Clouds. We then present the problem formulation and our optimal strategy of user scheduling based on Minimum First Derivative Length (MFDL) of load paths among the systems. Through theoretical analysis, we prove that our proposed algorithm Optimal user Scheduling for Cloud-of-Clouds (OSCC) can achieve the optimal solution.

Ever-increasing big data forces enterprises to migrate data to cloud storage systems. Data retrieval time from the cloud will directly affect the overall application performance. Meanwhile, sensitive data stored on cloud necessitates a robust security arrangement against cyberattacks. Therefore, it is imperative that both data retrieval time and data security should be taken into account simultaneously when designing a data placement strategy. In this paper, we formulate, design and evaluate the performance of a multi-objective evolutionary algorithm based data placement strategy. We show that our strategy offers users a choice to strike a balance between retrieval time and security through a set of uniformly distributed Pareto-optimal solutions. We evaluate and quantify the performance of our strategy on different cloud storage systems.

We study the problem of dynamically migrating a service in the cloud to satisfy an online sequence of mobile batch-request demands in a cost-effective way. The service may have single or multiple replicas, each running on a virtual machine. As the origin of mobile accesses frequently changes over time, this problem is particularly important for time-bounded services to achieve enhanced Quality of Service and cost effectiveness. Moving the service closer to the client locations not only reduces the service access latency but also minimizes the network costs for service providers. However, these benefits are not free. The migration comes at a cost of bulk-data transfer and service disruption, and hence, increasing the overall service costs. To gain the benefits of service migration while minimizing the caused monetary costs, we propose an efficient search-based algorithm Dmig to migrate a single server, and then extend it as a scalable algorithm, called mDmig , to the multi-server situation, a more general case in the cloud. Both algorithms are fully distributed, symmetric, and characterized by the effective use of historical access information to conduct virtual migration so that the limitations of local search in the cost reduction can be overcome. To evaluate the algorithms, we compared them with some existing algorithms and an off-line algorithm. Our simulation results showed that the proposed algorithms exhibit better performance in service migration by adapting to the changes of mobile access patterns in a cost-effective way.

This paper studies a scheduling problem of task graphs on a nondedicated networked computing platform. The networked platform is characterized by a set of fully connected processors such as a multiprocessor system that can be shared by multiple tasks. Therefore, the computation and communication capacities of the computing platform dynamically fluctuate. To deal with this fluctuations for high performance task graph computing, we propose an online dynamic resilience scheduling algorithm called Adaptive Scheduling Algorithm (ASA) that bears certain distinct features compared to existing algorithms. First, the proposed algorithm deliberately assigns tasks to idle processors in multiple rounds to prevent any unfavorable decisions and also to avoid inefficient assignments of certain key tasks to slow processors. Second, the algorithm adopts task duplication as an attempt to minimize serious increase of schedule length due to unexpected processor slowdown. Finally, a look-ahead message transmission policy is applied to save communication time and further improve the overall performance. Performance evaluation results are presented to demonstrate the effectiveness and competitiveness of our approaches when compared with the existing algorithms.

In recent years, recommendation systems have seen significant evolution in the field of knowledge engineering. Most of the existing recommendation systems based their models on collaborative filtering approaches that make them simple to implement. However, performance of most of the existing collaborative filtering-based recommendation system suffers due to the challenges, such as: (a) cold start, (b) data sparseness, and (c) scalability. Moreover, recommendation problem is often characterized by the presence of many conflicting objectives or decision variables, such as users' preferences and venue closeness. In this paper, we proposed MobiContext, a hybrid cloud-based bi-objective recommendation framework (BORF) for mobile social networks. The MobiContext utilizes multi-objective optimization techniques to generate personalized recommendations. To address the issues pertaining to cold start and data sparseness, the BORF performs data preprocessing by using the Hub-Average (HA) inference model. Moreover, the Weighted Sum Approach (WSA) is implemented for scalar optimization and an evolutionary algorithm (NSGA-II) is applied for vector optimization to provide optimal suggestions to the users about a venue. The results of comprehensive experiments on a large-scale real dataset confirm the accuracy of the proposed recommendation framework.

Multi-installment scheduling (MIS) has shown great effectiveness in minimizing the processing time for large-scale partitionable workloads. To derive an optimal MIS strategy, one has to explicitly determine optimal numbers of installments and processors. Existing studies tend to solve this problem by treating the influence of number of installments (and processors) w.r.t processing time as time-continuous functions and taking the derivative of these functions to determine the optimal values, which may lead to invalid solutions. In this paper, we employ periodic multi-installment scheduling (P-MIS) models for homogeneous and heterogeneous single-level tree networks. Using these models we make the following significant contributions. First, we derive a closed-form solution for an optimal number of installments based on a given network size and a fixed load distribution sequence. Second, we propose a heuristic algorithm for determining an optimal number of processors by first proving several important intermediate lemmas and theorems. Third, for heterogeneous systems, we propose a genetic algorithm to determine an optimal load distribution sequence. Finally, we conduct various experiments to illustrate the effectiveness of the proposed algorithms and perform rigorous analysis on the influence of load distribution sequence on processing time, on the basis of which a practical advice for determining a near-optimal load distribution sequence is given.

To respond to emergencies in a fast and an effective manner, it is of critical importance to have efficient evacuation plans that lead to minimum road congestions. Although emergency evacuation systems have been studied in the past, the existing approaches, mostly based on multi-objective optimizations, are not scalable enough when involve numerous time varying parameters, such as traffic volume, safety status, and weather conditions. In this paper, we propose a scalable emergency evacuation service, termed the MacroServ that recommends the evacuees with the most preferred routes towards safe locations during a disaster. Unlike many existing approaches that model systems with static network characteristics, our approach considers real-time road conditions to compute the maximum flow capacity of routes in the transportation network. The evacuees are directed towards those routes that are safe and have least congestion resulting in decreased evacuation time. We utilized probability distributions to model the real-life stochastic behaviors of evacuees during emergency scenarios. The results indicate that recommendation of appropriate routes during emergency scenarios play a critical role in quicker and safe evacuation of the population.

Mobile crowdsensing leverages mobile devices (e.g., smart phones) and human mobility for pervasive information exploration and collection; it has been deemed as a promising paradigm that will revolutionize various research and application domains. Unfortunately, the practicality of mobile crowdsensing can be crippled due to the lack of incentive mechanisms that stimulate human participation. In this paper, we study incentive mechanisms for a novel Mobile Crowdsensing Scheduling (MCS) problem, where a mobile crowdsensing application owner announces a set of sensing tasks, then human users (carrying mobile devices) compete for the tasks based on their respective sensing costs and available time periods, and finally the owner schedules as well as pays the users to maximize its own sensing revenue under a certain budget. We prove that the MCS problem is NP-hard and propose polynomial-time approximation mechanisms for it. We also show that our approximation mechanisms (including both offline and online versions) achieve desirable game-theoretic properties, namely truthfulness and individual rationality, as well as O(1) performance ratios. Finally, we conduct extensive simulations to demonstrate the correctness and effectiveness of our approach.

Lifetime reliability is an emerging concern in multiprocessor systems as escalating power density and hence temperature variation continues to accelerate wear-out leading to a growing prominence of device defects. In this paper, we propose a system-level approach that involves performance-aware mapping of multimedia applications on a multiprocessor system to jointly minimize energy consumption and temperature related wear-out. Fundamental to this approach is a simplified temperature model that incorporates not only the transient and the steady-state behavior (temporal effect), but also the temperature dependency on the surrounding cores (spatial effect). This model is validated against the temperature obtained using theHotSpottool with transient and steady-state simulations, and is shown to be accurate within 5.5°C, leading to an MTTF estimation accuracy of an average 21 percent with respect to the state-of-the-art approaches. The proposed temperature model is integrated in a gradient-based fast heuristic that controls the voltage and frequency of the cores to limit the average and peak temperature leading to a longer lifetime, simultaneously minimizing the energy consumption. Lifetime computation considers task remapping, which is a common feature available in modern multiprocessor systems. A linear programming approach is then proposed to distribute the cores of a multiprocessor system among concurrent applications to maximize the lifetime. Experiments conducted with a set of synthetic and real-life applications represented as synchronous data flow graphs demonstrate that the proposed approach minimizes energy consumption by an average 24 percent with 47 percent increase in lifetime. For concurrent applications, the proposed lifetime-aware core distribution results in an average 10 percent improvement in lifetime as compared to performance-based core distribution.

Anomaly detection from ECG stream is a key step leading to a significant success of the remote and auto-triggered cardiac event monitoring system. This effort requires an online processing and efficient analysis on the real-time data. Moreover, its computational complexity should be kept low so that the detection algorithm can be implemented even on a small computing device used in sensor network. In this paper, we present a novel fast and effective approach to identify abnormalities based on differences of heart beat morphologies. Our approach is inspired from time-series data mining techniques and statistical outlier detection methods. The experimental results overall (open public QT database) demonstrate high quality performance. In particular, it obtains 0.971, 0.995 and 0.994, on an average, for of sensitivity, specificity and accuracy for the respective performance metrics.

Generally, a parallel application consists of precedence constrained stochastic tasks, where task processing times and intertask communication times are random variables following certain probability distributions. Scheduling such precedence constrained stochastic tasks with communication times on a heterogeneous cluster system with processors of different computing capabilities to minimize a parallel application’s expected completion time is an important but very difficult problem in parallel and distributed computing. In this paper, we present a model of scheduling stochastic parallel applications on heterogeneous cluster systems. We discuss stochastic scheduling attributes and methods to deal with various random variables in scheduling stochastic tasks. We prove that the expected makespan of scheduling stochastic tasks is greater than or equal to the makespan of scheduling deterministic tasks, where all processing times and communication times are replaced by their expected values. To solve the problem of scheduling precedence constrained stochastic tasks efficiently and effectively, we propose a stochastic dynamic level scheduling (SDLS) algorithm, which is based on stochastic bottom levels and stochastic dynamic levels. Our rigorous performance evaluation results clearly demonstrate that the proposed stochastic task scheduling algorithm significantly outperforms existing algorithms in terms of makespan, speedup, and makespan standard deviation.

The articles in this special section focus on the economics and market mechanisms for cloud computing applications.

Providing scalable video streaming services for heterogeneous users in dynamic networked environments requires efficient and adaptive quality management mechanisms which deliver quality-customized services according to the client's preferences and adapt the services to cope with various network conditions. In this paper, we address the issue of quality adaptation for providing personalized scalable media streaming services in dynamic network environments. We propose a differentiated adaptive quality optimization algorithm, called Scalable Video Coding Quality Adaptation algorithm (SVC-QA), which adapts streaming quality based on both system-level and client-level optimization to optimize streaming quality according to network bandwidth conditions, content characteristics, a user's quality preferences, and buffering capacities of different client devices (e.g., mobile phones, PCs, HDTVs, etc.). Comparative studies are conducted to compare our proposed algorithms with other adaptive methods. We show that two-level SVC quality adaptation method can achieve better SVC streaming quality with both high peak signal-to-noise ratio (PSNR) and low quality variance under dynamic resource constraints. Moreover, the proposed distributed method reduces the computational complexities at the server side substantially, making it practical and flexible for providing scalable streaming services.

Cardiovascular disease is a class of diseases referring to functional abnormality of heart, which is the leading cause of deaths worldwide. Nowadays, one of the most popular methods to diagnose heart disease is based on electrocardiography (ECG) - a recording of the electrical activity of the heart. According to its characteristics, different patterns have been carefully studied in order to produce clinical decision-making. However, the diversity of patterns raises a lot of difficulties to memorize all of them. On the other hand, the historical patterns of prior ECG is also important to the diagnostic process. Therefore, in this research, we make an important use-inspired novel contribution namely, a framework based on semantic technologies, that allows to: (i) store ECG's features in a scalable linked database; (ii) flexibly access to ECG's characteristics through pattern queries; (iii) easily translate clinical ECG interpretation and heart disease diagnosis guides into semantic rules that can be automatically performed. Moreover, the new rules can be also easily and incrementally added to the framework.

With the emergence of information technologies, an overwhelming amount of data and information is generated everyday. Storing and processing this huge volume of data is named by a ubiquitous term: big data management. Cloud storage systems enhance reliability and availability of data by introducing redundancy, i.e., data replication, in the system, thereby protecting the data integrity from node failures which occur frequently in any large-scale storage system. However, efficiently determining the level of redundancy, i.e., number of data replicas, is not a trivial task for a cloud service provider (CSP). Traditional methods, which use a fixed number of replicas for all users regardless of the user's budget, do not achieve efficiency in terms of financial benefit of CSPs. This paper presents an efficient replication scheme that allows a CSP to determine the optimal number of replicas for each user depending on the user's budgetary constraint and the CSP's resource capacity while maximizing the financial benefit of the CSP. Numerical simulations were performed to assess the validity of our approach. The results show the scalability of the proposed scheme which can apply to real systems with an arbitrary number of users.

The thermal profile of multicore systems vary both within an application's execution (intra) and also when the system switches from one application to another (inter). In this paper, we propose an adaptive thermal management approach to improve the lifetime reliability of multicore systems by considering both inter- and intra-application thermal variations. Fundamental to this approach is a reinforcement learning algorithm, which learns the relationship between the mapping of threads to cores, the frequency of a core and its temperature (sampled from on-board thermal sensors). Action is provided by overriding the operating system's mapping decisions using affinity masks and dynamically changing CPU frequency using in-kernel governors. Lifetime improvement is achieved by controlling not only the peak and average temperatures but also thermal cycling, which is an emerging wear-out concern in modern systems. The proposed approach is validated experimentally using an Intel quad-core platform executing a diverse set of multimedia benchmarks. Results demonstrate that the proposed approach minimizes average temperature, peak temperature and thermal cycling, improving the mean-time-to-failure (MTTF) by an average of 2x for intra-application and 3x for inter-application scenarios when compared to existing thermal management techniques. Furthermore, the dynamic and static energy consumption are also reduced by an average 10% and 11% respectively.

This paper proposes a design-time (offline) analysis technique to determine application task mapping and scheduling on a multiprocessor system and the voltage and frequency levels of all cores (offline DVFS) that minimize application computation and communication energy, simultaneously minimizing processor aging. The proposed technique incorporates (1) the effect of the voltage and frequency on the temperature of a core; (2) the effect of neighboring cores' voltage and frequency on the temperature (spatial effect); (3) pipelined execution and cyclic dependencies among tasks; and (4) the communication energy component which often constitutes a significant fraction of the total energy for multimedia applications. The temperature model proposed here can be easily integrated in the design space exploration for multiprocessor systems. Experiments conducted with MPEG-4 decoder on a real system demonstrate that the temperature using the proposed model is within 5% of the actual temperature clearly demonstrating its accuracy. Further, the overall optimization technique achieves 40% savings in energy consumption with 6% increase in system lifetime.

Energy and reliability optimization are two of the most critical objectives for the synthesis of multiprocessor systems-on-chip (MPSoCs). Task mapping has shown significant promise as a low cost solution in achieving these objectives as standalone or in tandem as well. This paper proposes a multi-objective design space exploration to determine the mapping of tasks of an application on a multiprocessor system and voltage/frequency level of each tasks (exploiting the DVFS capabilities of modern processors) such that the reliability of the platform is improved while fulfilling the energy budget and the performance constraint set by system designers. In this respect, the reliability of a given MPSoC platform incorporates not only the impact of voltage and frequency on the aging of the processors (wear-out effect) but also on the susceptibility to soft-errors - a joint consideration missing in all existing works in this domain. Further, the proposed exploration also incorporates soft-error tolerance by selective replication of tasks, making the proposed approach an interesting blend of reactive and proactive fault-tolerance. The combined objective of minimizing core aging together with the susceptibility to transient faults under a given performance/energy budget is solved by using a multi-objective genetic algorithm exploiting tasks' mapping, DVFS and selective replication as tuning knobs. Experiments conducted with reallife and synthetic application graphs clearly demonstrate the advantage of the proposed approach.

Managing resources and cleverly pricing them on computing systems is a challenging task. Resource sharing demands careful load balancing and often strives to achieve a win-win situation between resource providers and users. Toward this goal, we consider a joint treatment of load balancing and pricing. We do not assume static pricing to determine load balancing, or vice versa. Instead, we study the relationship between the price that a computing node is charged and the load and revenue that it receives. We find that there exists an optimal price which maximizes the revenue. We then consider a multi-user environment and explore how the load from a user can be balanced on processors with existing loads. Finally, we derive an optimal price that maximizes the revenue in the multi-user environment. We evaluate the performance of the proposed algorithms through simulations.

The scheduling of tasks with deadlines on clusters is a key issue for offering quality-of-service (QoS) assurance. A critical challenge in real-time task scheduling is to handle various types of applications. This paper investigates the scheduling problem for processing a set of tasks comprising both divisible and indivisible real-time tasks on cluster systems. Indivisible tasks are characterized by the property that they need to be processed on their entirety on a single processor while divisible tasks can be distributed across several processing nodes by exploiting the underlying data parallelism. We propose a dynamic (on-line) real-time scheduling algorithm referred to as Hybrid Loads Push-Pull Scheduling (HLPPS) algorithm for handling a set of tasks comprising both divisible and indivisible real-time tasks on cluster systems. HLPPS is shown to efficiently exploit the parallelism in divisible tasks without undermining the schedulability of indivisible tasks and thereby optimize the overall performance. We consider two distinct network platforms - tightly coupled and loosely coupled clusters in designing the strategy. We conduct extensive performance evaluation studies to quantify the performance of the proposed algorithm under a variety of scenarios.

10.1109/TC.2014.3

Task mapping and scheduling are critical in minimizing energy consumption while satisfying the performance requirement of applications enabled on heterogeneous multiprocessor systems. An area of growing concern for modern multiprocessor systems is the increase in the failure probability of one or more component processors. This is especially critical for applications where performance degradation (e.g., throughput) directly impacts the quality of service requirement. This article proposes a design-time (offline) multi-criterion optimization technique for application mapping on embedded multiprocessor systems to minimize energy consumption for all processor fault-scenarios. A scheduling technique is then proposed based on self-timed execution to minimize the schedule storage and construction overhead at runtime. Experiments conducted with synthetic and real applications from streaming and nonstreaming domains on heterogeneous MPSoCs demonstrate that the proposed technique minimizes energy consumption by 22% and design space exploration time by 100x, while satisfying the throughput requirement for all processor fault-scenarios. For scalable throughput applications, the proposed technique achieves 30% better throughput per unit energy, compared to the existing techniques. Additionally, the self-timed execution-based scheduling technique minimizes schedule construction time by 95% and storage overhead by 92%.

Content distribution networks (CDNs) built on clouds have recently started to emerge. Compared to conventional CDNs, cloud-based CDNs have the benefit of cost efficient hosting services without owning infrastructure. However, resource provisioning and replica placement in cloud CDNs involve a number of challenging issues, mainly due to the dynamic nature of demand patterns. To deal with this dynamic nature, this paper proposes a set of novel algorithms to solve the joint problem of resource provisioning and caching (i.e., replica placement) for cloud-based CDNs with an emphasis on handling the dynamic demand patterns. Firstly, we propose a provisioning and caching algorithm framework called Differential Provisioning and Caching (DPC) algorithm, which aims to rent cloud resources to build CDNs and whereby to cache contents so that the total rental cost can be minimized while all demands are served. DPC consists of 2 steps. Step 1 first maximizes total demands supported by unexpired resources. Then, step 2 minimizes the total rental cost for new resources to serve all remaining demands. For each step we design both greedy and iterative heuristics, each with different advantages over the existing approaches. Moreover, to dynamically adjusts the placement of contents and route maps, we further propose the Caching and Request Balancing (CRB) algorithm, which is light-weight and thus can be frequently executed as a companion of DPC to maximize the total demands. Performance evaluation results are presented to demonstrate the effectiveness and competitiveness of our approaches when compared to existing algorithms.

Homogeneous multiprocessor systems with reconfigurable area (also known as Reconfigurable Multiprocessor Systems) are emerging as a popular design choice in current and future technology nodes to meet the heterogeneous computing demand of a multitude of applications enabled on these platforms. Application specific mapping decisions on such a platform involve partitioning a given application into software tasks (executed on one or more of the general purpose processors, GPPs) and the hardware tasks (realized as dedicated hardware on the reconfigurable area) to optimize and/or satisfy design constraints such as reliability, performance and design cost. Improving the reliability considering transient faults by increasing the number of checkpoints negatively impacts the reliability considering permanent faults. This trade-off is ignored in all prior studies on task mapping and scheduling. This paper proposes an optimization technique to decide the optimal number of checkpoints for the software tasks which minimizes aging of the GPPs while maximizing the transient fault-tolerance of the overall platform (GPPs and the reconfigurable area) and satisfying design cost and performance. Experiments conducted with synthetic and real-life application task graphs (cyclic and acyclic) demonstrate that the proposed technique minimizes aging and improves the platform lifetime by an average 60% as compared to the existing transient fault-aware techniques. Further, a gradient-based heuristic is proposed to minimize the design space exploration time by upto 500× with less than 5% deviation from optimal solution.

Shrinking transistor geometries, aggressive voltage scaling and higher operating frequencies have negatively impacted the lifetime reliability of embedded multi-core systems. In this paper, a convex optimization-based task-mapping technique is proposed to extend the lifetime of a multiprocessor systems-on-chip (MPSoCs). The proposed technique generates mappings for every application enabled on the platform with variable number of cores. Based on these results, a novel 3D-optimization technique is developed to distribute the cores of an MPSoC among multiple applications enabled simultaneously. Additionally, reliability of the underlying network-on-chip links is also addressed by incorporating aging of links in the objective function. Our formulations are developed for directed acyclic graphs (DAGs) and synchronous dataflow graphs (SDFGs), making our approach applicable for streaming as well as non-streaming applications. Experiments conducted with synthetic and real-life application graphs demonstrate that the proposed approach extends the lifetime of an MPSoC by more than 30% when applications are enabled individually as well as in tandem.

Shrinking transistor geometries, aggressive voltage scaling and higher operating frequencies have negatively impacted the dependability of embedded multicore systems. Most existing research works on fault-tolerance have focused on transient and permanent faults of cores. Intermittent faults are a separate class of defects resulting from on-chip temperature, pressure and voltage variations and lasting for a few cycles to several seconds or more. Operations of cores impacted by intermittent faults are suspended during these cycles but come back alive when conditions become favorable. This paper proposes a technique to model the availability of multiprocessor systems-on-chip (MPSoCs) with intermittent and reparable device defects. This model is based on Markov chain with stochastic fault distribution and can be applied even for permanent faults. Based on this model, a design space pruning technique is proposed to select a set of task mappings (with variable resource usage), which minimizes the task communication energy while satisfying the MPSoC availability constraint. Moreover, task migration overhead is also minimized, which is an important consideration for frequently occurring intermittent and temperature related faults, where prolonged system downtime during task re-mapping is not desired. Experiments conducted with real-life and synthetic application task graphs demonstrate that the proposed technique minimizes communication energy by 30% and reduces migration overhead by 50% as compared to the existing approaches.

Grids have been extensively deployed to handle various scientific and engineering applications that can be structured as bag-of-tasks (BoT). The scheduling of BoT applications on Grids is an important issue for achieving high performance. Grid scheduling involves a number of challenging issues, mainly due to the dynamic nature of the Grid. To deal with this dynamic nature, in this paper, we propose an online scheduling algorithm called prudent algorithm with replication (PAR) for scheduling Grid applications. PAR is shown to prudently make scheduling decisions in such a way that it can tolerate inaccurate performance predictions. Another point to note is that PAR adopts task duplication as an attempt to reduce serious schedule increases. Moreover, since the applications to be performed may widely vary in terms of their required hardware and software, we also capture the loads' various processing requirements in our algorithms, a unique feature that is applicable for running proprietary applications only on certain eligible processing nodes. Thus, in our problem formulation each application can only be processed by certain processors as both the applications and processing nodes are heterogeneous. We then present a task selection policy, referred to as requirement-aware load selection (RALS) policy to handle the contention of multiple applications that have various processing requirements but share the same computing resources. Based on RALS and PAR, we develop two scheduling algorithms: requirement-aware prudent algorithm with replication (RAPAR), and requirement-aware knowledge-free algorithm with replication (RAKAR). RAPAR and RAKAR address the scheduling of multiple BoT applications with heterogeneous processing requirements on Grids. RAPAR works in scenarios where inaccurate performance prediction information is provided whereas RAKAR works without any prediction information. Performance evaluation results are presented to demonstrate the effectiveness and competitiveness of our approaches when compared to existing algorithms.

While quality-adaptable applications are gaining increased popularity on embedded systems (especially multimedia applications), efficient scheduling techniques are necessary to explore this feature to achieve the optimal quality output. In addition to conventional real-time requirements, emerging challenges such as leakage power and multiprocessors further complicate the formulation and solution of adaptive application scheduling problems. In this paper, we propose a dynamic adaptive application scheduling scheme that efficiently distributes the runtime slack to achieve maximized execution quality under timing and dynamic/leakage energy constraints. Our proposed methods are threefold: First, for each task in the slack receiver group, a heuristic guided-search algorithm is proposed to select the optimal processor frequency to maximize the application execution quality. Second, we present an efficient slack receiver selection methodology aiming at identifying optimal slack receivers for quality maximization. Third, our framework is further extended to consider constraints brought by interprocessor communications, where we study the effects of slack inaccuracies introduced by transmission variations, and propose a local scaling approach to compensate the induced quality loss. Experimental results on synthesized tasks and a JPEG2000 codec show that the guided-search algorithm, aided by slack receiver selection, effectively outperforms contemporary approaches with at most 88 percent more quality improvement, whereas the local scaling contributes as large as 16.9 percent on top of the guided-search results.

In this paper, we study the strategies for efficiently achieving data staging and caching on a set of vantage sites in a cloud system with a minimum cost. Unlike the traditional research, we do not intend to identify the access patterns to facilitate the future requests. Instead, with such a kind of information presumably known in advance, our goal is to efficiently stage the shared data items to predetermined sites at advocated time instants to align with the patterns while minimizing the monetary costs for caching and transmitting the requested data items. To this end, we follow the cost and network models in [1] and extend the analysis to multiple data items, each with single or multiple copies. Our results show that under homogeneous cost model, when the ratio of transmission cost and caching cost is low, a single copy of each data item can efficiently serve all the user requests. While in multicopy situation, we also consider the tradeoff between the transmission cost and caching cost by controlling the upper bounds of transmissions and copies. The upper bound can be given either on per-item basis or on all-item basis. We present efficient optimal solutions based on dynamic programming techniques to all these cases provided that the upper bound is polynomially bounded by the number of service requests and the number of distinct data items. In addition to the homogeneous cost model, we also briefly discuss this problem under a heterogeneous cost model with some simple yet practical restrictions and present a 2-approximation algorithm to the general case. We validate our findings by implementing a data staging solver, whereby conducting extensive simulation studies on the behaviors of the algorithms.

Traditionally, the "best effort, cost free" model of Supercomputers/Grids does not consider pricing. Clouds have progressed towards a service-oriented paradigm that enables a new way of service provisioning based on "pay-as-you-go" model. Large scale many-task workflow (MTW) may be suited for execution on Clouds due to its scale-* requirement (scale up, scale out, and scale down). In the context of scheduling, MTW execution cost must be considered based on users' budget constraints. In this paper, we address the problem of scheduling MTW on Clouds and present a budget-conscious scheduling algorithm, referred to as ScaleStar (or Scale-*). ScaleStar assigns the selected task to a virtual machine with higher comparative advantage which effectively balances the execution time-and-monetary cost goals. In addition, according to the actual charging model, an adjustment policy, refer to as DeSlack, is proposed to remove part of slack without adversely affecting the overall makespan and the total monetary cost. We evaluate ScaleStar with an extensive set of simulations and compare with the most popular HEFT-based LOSS3 algorithm and demonstrate the superior performance of ScaleStar.

Shrinking transistor geometries, aggressive voltage scaling and higher operating frequencies have negatively impacted the dependability of embedded multiprocessor systems-on-chip (MPSoCs). Fault-tolerance and energy efficiency are the two most desired features of modern-day MPSoCs. For most of the multimedia applications, task communication energy constitutes more than 40% of the overall application energy. In this paper, an integer linear programming (ILP) based approach is proposed to reduce the communication energy and fault-tolerant migration overhead of throughput-constrained multimedia applications modeled using synchronous data flow graphs (SDFGs). The ILP is solved at compile-time for all fault-scenarios to generate task-core mappings satisfying an application throughput requirement. These mappings are stored in a table which is looked up at run-time as and when faults occur. Experiments conducted with real and synthetic applications demonstrate that the proposed technique reduces communication energy by an average 40% and migration overhead by 33% as compared to the existing fault-tolerant techniques.

The RAID6 provides high reliability using double-parity-update at cost of high write penalty. In this paper, we propose HRAID6ML, a new logging architecture for RAID6 systems for enhanced energy efficiency, performance and reliability. HRAID6ML explores a group of Solid State Drives (SSDs) and Hard Disk Drives (HDDs): Two HDDs (parity disks) and several SSDs form RAID6. The free space of the two parity disks is used as mirrored log region of the whole system to absorb writes. The mirrored logging policy helps to recover system from parity disk failure. Mirrored logging operation does not introduce noticeable performance overhead to the whole system. HRAID6ML eliminates the additional hardware and energy costs, potential single point of failure and performance bottleneck. Furthermore, HRAID6ML prolongs the lifecycle of the SSDs and improves the systems energy efficiency by reducing the SSDs write frequency. We have implemented proposed HRAID6ML. Extensive trace-driven evaluations demonstrate the advantages of the HRAID6ML system over both traditional SSD-based RAID6 system and HDD-based RAID6 system.

In many network applications, Bloom filters are used to support exact-matching membership query for their randomized space-efficient data structure with a small probability of false answers. In this paper, we extend the standard Bloom filter to Locality-Sensitive Bloom Filter (LSBF) to provide Approximate Membership Query (AMQ) service. We achieve this by replacing uniform and independent hash functions with locality-sensitive hash functions. Such replacement makes the storage in LSBF to be locality sensitive. Meanwhile, LSBF is space efficient and query responsive by employing the Bloom filter design. In the design of the LSBF structure, we propose a bit vector to reduce False Positives (FP). The bit vector can verify multiple attributes belonging to one member. We also use an active overflowed scheme to significantly decrease False Negatives (FN). Rigorous theoretical analysis (e.g., on FP, FN, and space overhead) shows that the design of LSBF is space compact and can provide accurate response to approximate membership queries. We have implemented LSBF in a real distributed system to perform extensive experiments using real-world traces. Experimental results show that LSBF, compared with a baseline approach and other state-of-the-art work in the literature (SmartStore and LSB-tree), takes less time to respond AMQ and consumes much less storage space.

Current FTL schemes have inevitable limitations in terms of memory requirement, performance, garbage collection overhead, and scalability. To overcome these limitations, we propose a workload adaptive flash translation layer referred to as WAFTL. WAFTL explores either page-level or block-level address mapping for normal data block based on access patterns. Page Mapping Block (PMB) is used to store random data and handle large number of partial updates. Block Mapping Block (BMB) is utilized to store sequential data and lower overall mapping table. PMB or BMB is allocated on demand and the number of PMB or BMB eventually depends on workload. An efficient address mapping is designed to reduce overall mapping table and quickly conduct address translation. WAFTL explores a small part of flash space as Buffer Zone to log writes sequentially and migrate data into BMB or PMB based on threshold. Static and dynamic threshold setting are proposed to balance performance and mapping table size. WAFTL has been extensively evaluated under various enterprise workloads. Benchmark results conclusively demonstrate that proposed WAFTL is workload adaptive and achieves up to 80% performance improvement, 83% garbage collection overhead reduction and 50% mapping table reduction compared to existing FTL schemes.

In the recent past, security-sensitive applications, such as electronic transaction processing systems, stock quote update systems, which require high quality of security to guarantee authentication, integrity, and confidentiality of information, have adopted heterogeneous distributed system (HDS) as their platforms. This is primarily due to the fact that single parallel-architecture-based systems may not be sufficient to exploit the available parallelism with the running applications. Most security-aware applications end up in handling dependence tasks, also referred to as Directed Acyclic Graph (DAG), on these HDSs. Unfortunately, most existing algorithms for scheduling such DAGs in HDS fail to fully consider security requirements. In this paper, we systematically design a security-driven scheduling architecture that can dynamically measure the trust level of each node in the system by using differential equations. To do so, we introduce task priority rank to estimate security overhead of such security-critical tasks. Furthermore, we propose a security-driven scheduling algorithm for DAGs which can achieve high quality of security for applications. Our rigorous performance evaluation study results clearly demonstrate that our proposed algorithm outperforms the existing scheduling algorithms in terms of minimizing the makespan, risk probability, and speedup. We also observe that the improvement obtained by our algorithm increases as the security-sensitive data of applications increases.

This paper investigates the problem of scheduling multiple divisible loads in networked computer systems with a particular emphasis in capturing two important real-life constraints, the arbitrary processor release times (or ready times) and heterogeneous processing requirements of different loads. We study two distinct cases of interest, static case, where processors' release times are predetermined and known, and dynamic case, where release times are unknown until processors are released. To address the two cases, we propose two novel scheduling strategies, referred to as Static Scheduling Strategy (SSS) and Dynamic Scheduling Strategy (DSS), respectively. In addition, we capture a task's processing requirements in our strategies, a unique feature that is applicable for handling loads on networks that run proprietary applications only on certain nodes. Thus, each task can only be processed by some certain nodes in our formulation. To handle the contention of multiple applications that have various processing requirements but share the same processing nodes, we propose an efficient load selection policy, referred to as Most Remaining Load First (MRF). We integrate MRF into SSS and DSS to address the problem of scheduling multiple divisible loads with arbitrary processor release times and heterogeneous requirements. We evaluate the strategies using extensive simulation experiments.

Data replication has been widely used as a mean of increasing the data availability of large-scale cloud storage systems where failures are normal. Aiming to provide cost-effective availability, and improve performance and load-balancing of cloud storage, this paper presents a cost-effective dynamic replication management scheme referred to as CDRM. A novel model is proposed to capture the relationship between availability and replica number. CDRM leverages this model to calculate and maintain minimal replica number for a given availability requirement. Replica placement is based on capacity and blocking probability of data nodes. By adjusting replica number and location according to workload changing and node capacity, CDRM can dynamically redistribute workloads among data nodes in the heterogeneous cloud. We implemented CDRM in Hadoop Distributed File System (HDFS) and experiment results conclusively demonstrate that our CDRM is cost effective and outperforms default replication management of HDFS in terms of performance and load balancing for large-scale cloud storage.

While performance-adaptable applications are gaining increased popularity on embedded systems (especially multimedia applications), efficient scheduling methods are necessary to explore such feature to achieve the most performance outcome. In addition to conventional scheduling requirements such as real-time and dynamic power, emerging challenges such as leakage power and multiprocessors further complicate the formulation and solution of adaptive application scheduling problems. In this paper, we propose a runtime adaptive application scheduling scheme that efficiently distributes the runtime slack in a task graph, to achieve maximized performance under timing and dynamic/leakage energy constraints. A guided-search heuristics is proposed to select the best-fit frequency levels that maximize the additional program cycles of adaptive tasks. Moreover, we devise a two-stage receiver task selection method that runs efficiently at runtime, in order to quickly find the slack distribution targets. Experiments on synthesized tasks and a JPEG2000 decoder are conducted to justify our approach. Results show that our method achieves at least 25% runtime performance increase compared to contemporary approaches, incurring negligible runtime overhead.

Combined computation and communication workload mapping and scheduling pose a major challenge in embedded NoC-based MPSoC design. While contemporary researches largely focus on data locality-centric mapping methodologies, unawareness of transmission route and timing may negatively impact the mapping efficiency. In this paper, we develop a unified communication-aware NoC-based MPSoC mapping and scheduling algorithm, in which a list-scheduling method is used to map prioritized tasks to the best fit processor, based on a transmission route-aware cost function. Our algorithm is able to realize precise and predictable packet routing in the process of task mapping, and achieve shorter end-to-end application execution time. To evaluate our algorithm, we conduct experiments using three real applications on a simulated NoC-based MPSoC platform. Comparison results show that our algorithm can achieve greatly improved overall end-to-end time, and about 38.3% less transmission time on a 3×3 mesh structure.

Data replication has been widely used as a mean of increasing the data availability of large-scale storage systems where failures are normal. Aiming to provide cost-effective availability, and improve performance and load-balancing of large-scale storage cluster, this paper presents a dynamic replication management scheme referred to as DRM. A model is developed to express availability as function of replica number. Based on this model, minimal replica number to satisfy availability requirement can be determined. DRM further places these replicas among Object-Based Storage Devices (OSD) in a balance way, taking into account different capacity and blocking probability of each OSD in heterogeneous environment. Proposed DRM can dynamically redistribute workloads among OSD cluster by adjusting replica number and location according to workload changing and OSD capacity. Our experiment results conclusively demonstrate that DRM is reliable and can achieve a significant average response time, and load balancing for large-scale OSD cluster.

BACKGROUND: RNA structure prediction problem is a computationally complex task, especially with pseudo-knots. The problem is well-studied in existing literature and predominantly uses highly coupled Dynamic Programming (DP) solutions. The problem scale and complexity become embarrassingly humungous to handle as sequence size increases. This makes the case for parallelization. Parallelization can be achieved by way of networked platforms (clusters, grids, etc) as well as using modern day multi-core chips. METHODS: In this paper, we exploit the parallelism capabilities of the IBM Cell Broadband Engine to parallelize an existing Dynamic Programming (DP) algorithm for RNA secondary structure prediction. We design three different implementation strategies that exploit the inherent data, code and/or hybrid parallelism, referred to as C-Par, D-Par and H-Par, and analyze their performances. Our approach attempts to introduce parallelism in critical sections of the algorithm. We ran our experiments on SONY Play Station 3 (PS3), which is based on the IBM Cell chip. RESULTS: Our results suggest that introducing parallelism in DP algorithm allows it to easily handle longer sequences which otherwise would consume a large amount of time in single core computers. The results further demonstrate the speed-up gain achieved in exploiting the inherent parallelism in the problem and also elicits the advantages of using multi-core platforms towards designing more sophisticated methodologies for handling a fairly long sequence of RNA. CONCLUSION: The speed-up performance reported here is promising, especially when sequence length is long. To the best of our literature survey, the work reported in this paper is probably the first-of-its-kind to utilize the IBM Cell Broadband Engine (a heterogeneous multi-core chip) to implement a DP. The results also encourage using multi-core platforms towards designing more sophisticated methodologies for handling a fairly long sequence of RNA to predict its secondary structure.

Scheduling multisource divisible loads is a challenging task as different sources should cooperate and share their computing power with others to balance their loads and minimize total computational time. In this study, we attempt to address a generalized divisible load scheduling problem for handling loads from multiple sources on arbitrary networks. This problem is all the more challenging as 1) the topology is arbitrary, 2) in such networks, it is difficult to decide from which source and which route a processing node should receive loads, and 3) processing nodes must be allocated to different sources when they become available. We study two distinct cases of interest, static case and dynamic case, and propose two novel strategies, referred to as static scheduling strategy (SSS) and dynamic scheduling strategy (DSS), respectively. Both strategies work in an iterative fashion. In each iteration, they will use a novel graph partitioning (GP) scheme to partition the network such that each source in the network gains a portion of network resources and then these sources cooperate to process their loads. We analyze the performance of DSS using queuing theory and derive upper bounds on a load's average waiting time and a source's average queue length. We use simulation to verify the usefulness and effectiveness of SSS and DSS. Our findings reveal an interesting ¿load insensitive¿ propertyof SSS and also verify the theoretical upper bound of average queue length at each source in the dynamic case.

The interconnection networks used by current fine grain FPGAs are not scalable for very big array sizes. To address this issue, we apply the GALS (globally asynchronous and locally synchronous) paradigm to build scalable FPGAs. The logic resources are divided into locally synchronous tiles and asynchronous communications among different tiles. To route the asynchronous communications, we build a serial network-on-chip. Targeting streaming applications, we propose a design flow that maps user applications to our new FPGA architecture. To validate our architecture and design flow, we build an emulation prototype and develop a JPEG baseline encoder as the case study. We have successfully demonstrated the concept and predict a maximum frequency of 224 MHz for designs mapping to sFPGA2 architecture.

In this paper, we evaluate and compare the performance of several spanning tree routing strategies for divisible load scheduling on arbitrary graphs and derive recommendations as to which routing strategy provides a better trade-off between complexity and time performance. We consider a network comprising heterogeneous processors interconnected by heterogeneous links in an arbitrary manner. We evaluate the performance over a wide range of arbitrary dense graphs with varying connectivity and processor densities and study the effect of network scalability. In addition, we introduce a novel spanning tree routing strategy, which is referred to as minimum equivalent network spanning tree (EST), and analyze its performance. We apply the resource-aware optimal load distribution with optimal sequencing (RAOLD-OS) scheduling algorithm presented in the literature for obtaining an optimal solution. This study attempts to pool all known and applicable divisible load scheduling algorithms for arbitrary networks and presents a collective and comparative view of their performance.

With our society more information-driven, we have begun to distribute data in wide-area storage systems. At the same time, both physical failure and logic error have made it difficult to bring the necessary recovery to bear on remote data disaster, and understanding this proceeding. We describe ARRAY, a system architecture for data disaster recovery that combines reliability, storage space, and security to improve performance for data recovery applications. The paper presents an exhaustive analysis of the design space of ARRAY systems, focusing on the trade-offs between reliability, storage space, security, and performance that ARRAY must make. We present RSRAII (Replication-based Snapshot Redundant Array of Independent Imagefiles) which is a configurable RAID-like data erasure-coding, and also others benefits come from consolidation both erasure-coding and replication strategies. A novel algorithm is proposed to improve snapshot performance referred to as SMPDP (Snapshot based on Multi-Parallel Degree Pipeline).

Disk idle periods in server workload are short, which significantly limits the effectiveness of underline disk power management. To release this limitation, we present a cooperative power management (referred to as CPM) scheme to save energy with performance guarantee for object-based storage cluster. CPM reclaims idle memories of neighboring object-based storage devices (OSDs) over high speed network as remote cache to store evicted objects. Then requests missed in local cache could be hit by remote cache, and local disk does not necessarily spin back up to service these requests. Hence, CPM can artificially create long idle periods to provide more opportunities for underlying disk power management. CPM minimizes the risk of performance and energy penalty by spinning down disks only when predicted idle period is long enough to justify state-transition energy. Our rigorous experiment results conclusively demonstrate that CPM can dynamically adapt to workload changes and outperform existing solutions in terms of energy saving and performance for large-scale OSD cluster.

Fault-tolerant scheduling is an imperative step for large-scale computational grid systems, as often geographically distributed nodes co-operate to execute a task. By and large, primary-backup approach is a common methodology used for fault tolerance wherein each task has a primary copy and a backup copy on two different processors. In this paper, we identify two cases that may happen when scheduling dependent tasks with primary-backup approach. We derive two important constraints that must be satisfied. Further, we show that these two constraints play a crucial role in limiting the schedulability and overloading efficiency of backups of dependent tasks. We then propose two strategies to improve schedulability and overloading efficiency, respectively. We propose two algorithms (MRC-ECT and MCT-LRC), to schedule backups of independent jobs and dependent jobs, respectively. MRC-ECT is shown to guarantee an optimal backup schedule in terms of replication cost for an independent task, while MCT-LRC can schedule a backup of a dependent task with minimum completion time and less replication cost. We conduct extensive simulation experiments to quantify the performance of the proposed algorithms.

In this paper, we present two heuristic energy-aware scheduling algorithms (EGMS and EGMSIV) for scheduling task precedence graphs in an embedded multiprocessor system having processing elements with dynamic voltage scaling capabilities. Unlike most energy-aware scheduling algorithms that consider task ordering and voltage scaling separately from task mapping, our algorithms consider them in an integrated way. EGMS uses the concept of energy gradient to select tasks to be mapped onto new processors and voltage levels. EGM-SIV extends EGMS by introducing intra-task voltage scaling using a Linear Programming (LP) formulation to further reduce the energy consumption. Through rigorous simulations, we compare the performance of our proposed algorithms with a few approaches presented in the literature. The results demonstrate that our algorithms are capable of obtaining energy-efficient schedules using less optimization time. On the average, our algorithms produce schedules which consume 10% less energy with more than 47% reduction in optimization time when compared to a few approaches presented in the literature. In particular, our algorithms perform better in generating energy-efficient schedules for larger task graphs. Our results show a reduction of up to 57% in energy consumption for larger task graphs compared to other approaches.

In designing energy-aware CPU scheduling algorithms for real-time embedded systems, dynamic slack reclamation techniques significantly improve system Quality-of-Service (QoS) and energy efficiency. However, the limited schemes in this domain either demand high complexity or can only achieve limited QoS. In this paper, we present a novel low complexity runtime scheduling algorithm for the Imprecise Computation (IC) modeled tasks. The target is to maximize system QoS under energy constraints. Our proposed algorithm, named Gradient Curve Shifting (GCS), is able to decide the best allocation of slack cycles arising at runtime, with very low complexity. We study both linear and concave QoS functions associated with IC modelde tasks, on non-DVS and DVS processors. Furthermore, we apply the intea-task DVS technique to tasks and achieve as large as 18% more of the system QoS compared to the conventional “optimal” solution which is inter-task DVS based.

Object-based storage is emerging as a next generation of distributed storage technology. Aiming at improving the performance and load-balancing of large-scale object-based storage system, we present a dynamic weight-based cooperative caching scheme referred to as DWC2, which allows an object-based storage device (OSD) to use the available free cache of the neighbouring OSD. Our proposed DWC2replaces objects based on their weights which is a function of object size, popularity and replica number, and dynamically partitions the memory of OSD into local cache and remote cache according to activity workload. An object data is cached in local cache or remote cache of the cooperative OSDs, thus increasing cache hit ratio, reducing expensive disk access time as well as improving load balance. We benchmarked our proposed DWC2with existing cooperative caching schemes under various OSD environments. Our rigorous experiment results conclusively demonstrate that our DWC2is scalable and can achieve a significant cache hit ratio, average response time, and load balancing for large-scale OSD cluster.

Server-side Web caching is an important technique used to reduce the user perceived latency (UPL). In large-scale multimedia systems, there are many Web proxies, connected with a multimedia server, that can cache some most popular multimedia objects. Multimedia objects have some particular characteristics, e.g., strict QoS requirements. Hence, even some efficient conventional caching strategies based on cache hit ratio, meant for non-multimedia objects, will confront some problems in dealing with the multimedia objects. If we consider additional resources of proxy besides cache space, say bandwidth, we can readily observe that high hit ratio may deteriorate the entire system performance. In this paper, we propose a novel placement model for networked multimedia systems, referred to as Hk/T model, which considers the combined influence of arrival rate, size, and playback time to select the objects to be cached. Based on this model, we propose an innovative Web cache replacement algorithm, named as ART-greedy algorithm, which can balance the load among the proxies and achieve a minimum average response time (ART) of the requests. Using an event-driven simulation, we evaluate the performance of our proposed algorithm under several situations. Our experimental results conclusively demonstrate that ART-greedy algorithm outperforms the most popular and commonly used LFU (least frequently used) algorithm significantly.

In a large-scale multimedia storage system (LMSS) where client requests for different multimedia objects may have different demands, placement and replication of the objects is an important factor, as it may result in an imbalance in server loading across the system. Since replica management and load balancing is all the more a crucial issue in multimedia systems, in the literature this problem is handled by centralized servers. Each object storage server (OSS) responses the requests coming from the centralized servers independently and has no communication with other OSSs among the system. In this paper, we design a novel distributed load balancing strategy of LMSS, in which the OSSs can cooperate together to achieve a high performance. Such OSS modeled as an M/M/m system, can replicate the objects to and balance the requests among other servers to achieve an optimal average waiting time (AWT) of the requests in the system. We validate the performance of the system via rigorous simulations with respect to several influencing factors and prove that our proposed strategy is scalable, flexible and efficient for the real-life applications.

Interactions between transcription factors (TFs) are necessary for deciphering the complex mechanisms of transcription regulation in eukaryotes. In this paper, we proposed a novel HV-kernel based Support Vector Machine classifier (HV-SVM) to predict TF-TF interactions based on their protein domain information and GO annotations. Specifically, two types of pairwise kernels, namely, a horizontal kernel and a vertical kernel, were combined to evaluate the similarity between a pair of TFs, and a Genetic algorithm was used to obtain kernel and feature weights to optimize the classifier's performance. We applied our proposed HV-SVM method to predict TF interactions for Homo sapiens and Mus muculus. We obtained accuracy and F-measures of over 85% and an AUC of almost 93%, demonstrating that HV-SVM can accurately predict TF-TF interactions even in the higher and more complex eukaryotes.

10.1109/CCNC.2007.71

Fault-tolerant scheduling is an imperative step for large-scale computational Grid systems, as often geographically distributed nodes co-operate to execute a task. By and large, the primary-backup approach is a common methodology used for fault tolerance where in each task has a primary copy and a backup copy on two different processors. Backup overloading has been proposed to reduce replication cost by allowing the backup copy to overload with other backup copies on the same processor. In this paper, we consider two classes of independent tasks where in both the classes have fault-tolerance requirements. Furthermore, Class 1 tasks require the response time to be as short as possible when a fault occurs, while Class 2 tasks prefer backups with minimum replication cost. We propose two algorithms, called the MRC-ECT algorithm and the MCT-LRC algorithm. Algorithm MRC-ECT is shown to guarantee an optimal backup schedule in terms of replication cost, while MCT-LRCcan schedule a backup with minimum completion time and low replication cost. We conduct extensive simulation experiments to quantify the performance of the proposed algorithms.

In this paper, we propose novel low-energy static and dynamic scheduling algorithms with low computational complexities, for heterogeneous multiprocessor real-time embedded systems. We consider task graphs with deadlines and precedence relationships to satisfy. We propose a novel scheme, referred to as “critical-path information track-and update”, based on critical-path analysis to distribute the slack-time over tasks such that energy consumption is minimized, while guaranteeing the precedence and timing constraints. Our dynamic scheduling algorithm applies the static scheduling algorithm during runtime based on the updated average-case execution demands of tasks. Our simulation results show that the proposed static scheduling algorithm consumes only 2% of the computational time with no degradation in energy savings, whereas the dynamic scheduling algorithm delivers up to 25% more energy savings while reducing the computational time overhead by more than 90%, when compared with recent heterogeneous multiprocessor scheduling algorithms.

In this paper, we propose novel low-energy scheduling algorithms with low computational complexities for the heterogeneous body area network (BAN) systems, considering task graphs with deadlines (timing constraints) and precedence relationships to satisfy. Our proposed novel scheme, referred to as "critical-path information track-and-update", analyses the critical-paths, identifies the slack and distributes it over tasks such that the overall energy consumption is minimised. Our dynamic scheduling algorithm utilises the results from the static scheduling algorithm and attempts to aggressively reduce the energy consumption. Simulations for the task graph for a typical BAN application show that our static and dynamic scheduling algorithms deliver 25% and 15% more energy savings respectively compared to typical slack reclamation based scheduling algorithms.

In this paper, we propose load distribution strategies for divisible loads for networked computing environments where computation and communication resource characteristics are unknown and/or vary with time. The principle on which our strategies are formulated is based on using probing loads to estimate the network characteristics and using them to determine the best possible load distribution. This work extends the adaptive strategies proposed in an earlier work to multilevel general networks. These networks, while being more challenging, also offer several opportunities that can be exploited to make the probing phase more efficient. We propose two strategies, one static, which caters to the presence of unknown parameters, and the other dynamic, which caters to both unknown as well as time-varying network parameters. The proposed strategies are robust, resilient, and easily adaptable to network fluctuations. The algorithms are also shown to have a tracking ability, a property that is important in dynamic environments. Examples are presented to illustrate the salient features of these strategies.

Large-scale distributed applications are subject to frequent disruptions due to resource contention and failure. Such disruptions are inherently unpredictable and, therefore, robustness is a desirable property for the distributed operating environment. In this work, we describe and evaluate a robust topology for applications that operate on a spanning tree overlay network. Unlike previous work that is adaptive or reactive in nature, we take a proactive approach to robustness. The topology itself is able to simultaneously withstand disturbances and exhibit good performance. We present both centralized and distributed algorithms to construct the topology, and then demonstrate its effectiveness through analysis and simulation of two classes of distributed applications: Data collection in sensor networks and data dissemination in divisible load scheduling. The results show that our robust spanning trees achieve a desirable trade-off for two opposing metrics where traditional forms of spanning trees do not. In particular, the trees generated by our algorithms exhibit both resilience to data loss and low power consumption for sensor networks. When used as the overlay network for divisible load scheduling, they display both robustness to link congestion and low values for the makespan of the schedule

In this paper, we address several issues that are imperative to grid environments such as handling resource heterogeneity and sharing, communication latency, job migration from one site to other, and load balancing. We address these issues by proposing two job migration algorithms, which are MELISA (modified ELISA) and LBA (load balancing on arrival). The algorithms differ in the way load balancing is carried out and is shown to be efficient in minimizing the response time on large and small-scale heterogeneous grid environments, respectively. MELISA, which is applicable to large-scale systems (that is, interGrid), is a modified version of ELISA in which we consider the job migration cost, resource heterogeneity, and network heterogeneity when load balancing is considered. The LBA algorithm, which is applicable to small-scale systems (that is, intraGrid), performs load balancing by estimating the expected finish time of a job on buddy processors on each job arrival. Both algorithms estimate system parameters such as the job arrival rate, CPU processing rate, and load on the processor and balance the load by migrating jobs to buddy processors by taking into account the job transfer cost, resource heterogeneity, and network heterogeneity. We quantify the performance of our algorithms using several influencing parameters such as the job size, data transfer rate, status exchange period, and migration limit, and we discuss the implications of the performance and choice of our approaches.

In this paper, we propose distributed algorithms referred to as resource-aware dynamic incremental scheduling (RADIS) strategies. Our strategies are specifically designed to handle large volumes of computationally intensive arbitrarily divisible loads submitted for processing at cluster/grid systems involving multiple sources and sinks (processing nodes). We consider a real-life scenario, wherein the buffer space (memory) available at the sinks (required for holding and processing the loads) varies over time, and the loads have deadlines and propose efficient "pull-based" scheduling strategies with an admission control policy that ensures that the admitted loads are processed, satisfying their deadline requirements. The design of our proposed strategies adopts the divisible load paradigm, referred to as the divisible load theory (DLT), which is shown to be efficient in handling large volume loads. We demonstrate detailed workings of the proposed algorithms via a simulation study by using real-life parameters obtained from a major physics experiment.

In this paper, we propose a novel resource scheduling strategy, referred to as the Multi-Resource Scheduling (MRS) algorithm, which is capable of handling several resources to be used among jobs that arrive at a Grid Computing Environment. We propose a model in which the job and resource characteristics are captured together and are used in the scheduling strategy. To do so, we introduce the concept of virtual map and resource potential. Based on the proposed model, simulations with realistic workload traces were conducted to quantify the performance. We compare our strategy with some of the commonly used algorithms, and show that MRS renders a higher performance in all cases. Our experimental results clearly show that MRS outperforms other strategies and we highlight the impact and importance of our strategy.

In distributed multimedia storage systems where client requests for different multimedia objects may have different demands, placement and replication of the objects is an important factor, as it may result in an imbalance in server loading across the system. Replica management and load balancing is all the more a crucial issue for large-scale multimedia systems. We design and analyze a heuristic static strategy to determine the placement, number of replicas of the objects, and balance the client requests among the servers, to minimize the average waiting time (AWT) of the requests. Our strategy exploits clever virtual routing technique to strike a balance among the servers for load balancing. We validate the performance via rigorous simulations with respect to several influencing factors and compare with a system that uses hashing function only for load balancing.

In this paper, we proposed to setup a distributed multimedia system which aggregates the capacity of multiple servers to provide customized multimedia services in a cost-effective way. Such a system enables clients to customize their services by specifying the service delay or the viewing times. We developed an experimental prototype in which media servers can cooperate in streams caching, replication and distribution. We applied a variety of stream distribution algorithms to the system and studied their performance under the real-life situations with limited network resources and varying request arrival pattern. The results show such a system can provide cost-effective services and be applied to practical environments.

In this paper, we classify the dynamic distributed load balancing algorithms for heterogenous distributed computer systems into three policies: queue adjustment policy (QAP), rate adjustment policy (RAP), and queue and rate adjustment policy (QRAP). We propose two efficient algorithms, referred to as rate-based load balancing via virtual routing (RLBVR) and queue-based load balancing via virtual routing (QLBVR), which belong to the above RAP and QRAP policies, respectively. We also consider algorithms estimated load information scheduling algorithm (ELISA) and perfect information algorithm, which were introduced in the literature, to implement QAP policy. Our focus is to analyze and understand the behaviors of these algorithms in terms of their load balancing abilities under varying load conditions (light, moderate, or high) and the minimization of the mean response time of jobs. We compare the above classes of algorithms by a number of rigorous simulation experiments to elicit their behaviors under some influencing parameters, such as load on the system and status exchange intervals. We also extend our experimental verification to large scale cluster systems such as a mesh architecture, which is widely used in real-life situations. From these experiments, recommendations are drawn to prescribe the suitability of the algorithms under various situations

In this paper, we address the performance of distributed database systems with buffer constraints. Specifically, our objective is to design and analyze efficient data allocation and replication strategies to minimize the total servicing cost for an arbitrary read/write request sequence, under finite buffer constraints of the nodes in the system. When the available buffer space in a node is not enough to store a copy of an object, the decision has to be made on whether or not we should evict one or more objects in use to give room for the new object copy. In this paper, we design and analyze the data replication strategies with the model of dynamic window mechanism (DWM) algorithm jointly implemented with different types of object replacement strategies (no replacement, LRU, and LFU) commonly found in practice. We consider situations wherein the object sizes are identical as well as heterogeneous. We will show the impact on the performance of the allocation and replication strategies due to the limited local database buffer capacities. We analyze and quantify theoretically (using competitive analysis) the performances of all the proposed algorithms. Further, we perform rigorous simulation experiments to validate the findings with respect to several influencing parameters. Several useful conclusions are drawn based on the experimental results and we highlight the usefulness of the algorithms under different situations

In this paper, we present a generalized approach to retrieve a long-duration movie requested using a network-based video-on-demand service infrastructure employing multiple servers. We design and analyze a play-while-retrieve (PWR) playback strategy for this multiserver environment such that the access time (waiting time for the clients) is minimized. For this strategy, we use both the single-installment and multi-installment retrieval strategies to analyze the performance of the service system. For the above-mentioned retrieval strategies, we explicitly derive closed-form expressions for a minimum access time. For the case of multi-installment retrieval strategy, we conduct asymptotic performance analysis that quantifies the ultimate performance bounds of our strategy. We demonstrate analytically the impact of a large-scale network, as well as the impact of indefinitely increasing the number of installments, on the performance of such a multiserver service system. We then address the problem of buffer management at the client site, which is a closely related issue that has a significant influence on the performance of the strategy, and also serves as a key issue in making the service system attractive for clients. We derive relationships that quantify the minimum amount of buffer expected at the client site to have a smooth presentation with this multiserver service structure. Finally, we perform simulation experiments to verify all our theoretical findings. In the experiments, we compare the performance of PWR strategy with that of play-after-retrieve strategy, and discuss certain important points that are crucial for implementing a real-life working multiserver service system.

In this paper, we propose a novel strategy, referred to as window-assisted video partitioning (WAVP), for efficiently rendering network-based multimedia services within delay constraints. To minimize the service cost and maximize the number of requests that can be successfully served under resources constraints (cache capacity and link bandwidth), our WAVP strategy partitions videos into multiple portions and delivers them by adaptive schedule windows. Based on a mathematical analysis, it is shown that the service cost can be optimized by managing video portions with the schedule windows and it can improve resource utilization to partition video streams into multiple portions. We analyze the performance under several influencing parameters such as link availability, cache capacity, delay bound and partition gradients. Simulation results show that our proposed method can significantly reduce the service cost and achieve a high acceptance ratio under the constraints of network resources.

The processing time needed for motion estimation usually accounts for a significant part of the overall processing time of the video encoder. To improve the video encoding speed, reducing the execution time for motion estimation process is essential. Parallel implementation of video encoding systems using either the software or the hardware approach has attracted much attention in the area of real time video coding. In this paper, we attempt to implement a video encoder on a bus network. Usually, for such a parallel system, the key concern is associated with partitioning and balancing of the computational load among the processors such that the overall processing time of the video encoder is minimized. With the use of the divisible load theory (DLT) paradigm, a strip-wise load partitioning/balancing scheme, a load distribution strategy, two implementation strategies are developed to exploit the data parallelism inherent in the video encoding process. The striking feature of our design is that,both the granularity of the load partitions and all the associated overheads caused during parallel video encoding process can be explicitly considered. This significantly contributes to the minimization of the overall processing time of the video encoder. Extensive experimental studies are carried out to test the effectiveness of the proposed strategies. The performance of the parallel video encoder is quantified using the metrics speedup and performance gain, respectively. The experimental results show that our strategies are effective for exploiting the available parallelism inherent in the video encoding process and provide a theoretical insight on how to analytically quantify and minimize the overall processing time of a parallel system. The proposed strategies can be easily extended and applied to improve other existing parallel systems.

In this paper, we design a multiprocessor strategy that exploits the computational characteristics of the algorithms used for biological sequence comparison proposed in the literature. We employ divisible load theory (DLT) that is suitable for handling large scale processing on network based systems. For the first time in the domain of DLT, the problem of aligning biological sequences is attempted. The objective is to minimize the total processing time of the alignment process. In designing our strategy, DLT facilitates a clever partitioning of the entire computation process involved in such a way that the overall time consumed for aligning the sequences is a minimum. The partitioning takes into account the computation speeds of the nodes and the underlying communication network. Since this is a real-life application, the post-processing phase becomes important, and hence we consider propagating the results back in order to generate an exact alignment. We consider several cases in our analysis such as deriving closed-form solutions for the processing time for heterogeneous, homogeneous, and networks with slow links. Further, we attempt to employ a multiinstallment strategy to distribute the tasks such that a higher degree of parallelism can be achieved. For slow networks, our strategy recommends near-optimal solutions. We derive an important condition to identify such cases and propose two heuristic strategies. Also, our strategy can be extended for multisequence alignment by utilizing a clustering strategy such as the Berger-Munson algorithm proposed in the literature. Finally, we use real-life DNA samples of house mouse mitochondrion (Mus Musculus Mitochondrion, NC_001569) consisting of 16,295 residues and the DNA of human mitochondrion (Homo Sapiens Mitochondrion, NC_001807) consisting of 16,571 residues, obtainable from the GenBank, in our rigorous simulation experiments to illustrate all the theoretical findings.

Video-on-demand (VoD) applications place very high requirements on the delivery medium. High-quality services should provide for a timely delivery of the data-stream to the clients plus a minimum of playback disturbances. The major contributions of this paper are that it proposes a multiserver, multi-installment (MSMI) solution approach (sending the document in several installments from each server) to the delivery problem and achieves a minimization of the client waiting time, also referred to as the access time (AT) or start-up latency in the literature. By using multiple spatially distributed servers, we are able to exploit slow connections that would otherwise prevent the deployment of video-on-demand-like services, to offer such services in an optimal manner. Additionally, the delivery and playback schedule that is computed by our approach is loss-aware in the sense that it is flexible enough to accommodate packet losses without interrupts. The mathematical framework presented covers both computation and optimization problems associated with the delivery schedule, offering a complete set of guidelines for designing MSMI VoD services. The optimizations presented include the ordering of the servers and determining the number of installments based on the packet-loss probabilities of the communication links. Our analysis guarantees the validity of a delivery schedule recommended by the system by providing a percentage of confidence for an uninterrupted playback at the client site. This, in a way, quantifies the degree of quality of service rendered by the system and the MSMI strategy proposed. The paper is concluded by a rigorous simulation study that showcases the substantial advantages of the proposed approach and explores how optimization of the schedule parameters affects performance.

We designed and analyzed three variable bit rate caching strategies for continuous media data. Firstly, we use a just-in-time scheme to improve the utilization efficiency of the resource (the cache space and the cache I/O bandwidth). Secondly, the assumption of constant retrieval bandwidth in past caching algorithms is nearly impossible in a practical multimedia system. We propose the strategy used in the case of variable retrieval bandwidth as well as constant retrieval bandwidth. Thirdly, the phenomena of "switching" in past caching algorithms significantly impacts the system performance, therefore, we propose and satisfy a non-switch constraint, so that the probability of a switching operation is reduced to a maximum extent. The simulation result confirms our analysis.

We address the issue of minimizing the per user service cost and at the same time maximizing the number of requests that can be served by distributed video-on-reservation (VOR) systems. We propose a heuristic algorithm, destination-based stream scheduling (DBS) algorithm, which combines the concept of multicast routing and network caching with end-to-end delay constraints. Our simulation results show that our algorithm can reduce the service cost, balance the network load and achieve a high acceptance ratio.

The problem of distributing and processing a divisible load in a heterogeneous linear net-work of processors with arbitrary processors release times is considered. A divisible load is very large in size and has computationally intensive CPU requirements. Further, it has the property that the load can be partitioned arbitrarily into any number of portions and can be scheduled onto processors independently for computation. The load is assumed to arrive at one of the farthest end processors, referred to as boundary processors, for processing. The processors in the network are assumed to have non-zero release times, i.e., the time instants from which the processors are available for processing the divisible load. Our objective is to design a load distribution strategy by taking into account the release times of the processors in such a way that the entire processing time of the load is a minimum. We consider two generic cases in which all processors have identical release times and when all processors have arbitrary release times. We adopt both the single and multi-installment strategies pro-posed in the divisible load scheduling literature in our design of load distribution strategies, wherever necessary, to achieve a minimum processing time. Finally, when optimal strate-gies cannot be realized, we propose two heuristic strategies, one for the identical case, and the other for non-identical release times case, respectively. Several conditions are derived to determine whether or not optimal load distribution exists and illustrative examples are provided for the ease of understanding.

In this paper, we address the problem of minimizing the cost of transferring a document or a file requested by a set of users geographically separated on a network of nodes. We concentrate on theoretical aspects of data migration and caching on high-speed networks. Following the information caching paradigm introduced in the literature, we present polynomial time optimal caching strategies that minimize the total monetary cost of all the service requests by the users on a high-speed network. We consider a scenario in which a large pool of customers from one or more remote sites on a network demand a document, situated at some site, for their use. We also assume that the users can request the document at different time instants. This process of distributing the requested document incurs communication costs due to the use of communication resources and caching costs of the document at some server sites before it is delivered to the users at their desired time instances. We configure the network as a fully connected topology in which the service providers manage and control the distribution of the requested document among the users. For a high-speed network, we show that a single copy of the requested document is sufficient to serve all the user requests in an optimal manner. We extend the study to a homogeneous case in which the communication costs are identical and caching costs at all the sites are identical. In this case, we demonstrate the adaptability of the algorithm in generating more than one copy when needed by the minimization process. Using these strategies, the network service providers can decide when, where, and for how long the requested documents must be cached at vantage sites to obtain an optimal solution. Illustrative examples are provided to ease the understanding.

In this paper we address the problem of scheduling a large size divisible load on a hypercube cluster of processors. Unlike in earlier studies in the divisible load theory (DLT) literature, here, we assume that the processors have finite-size buffers. Further we impose constraints on the extent to which the load can be divided, referred to as granularity constraint. We first present the closed-form solutions for the case with infinite-size buffers. For the case with load granularity constraint, we propose a simple algorithm to find the sub-optimal solution and then we analyze the case when these buffer are of finite size. For this case, we present an elegant strategy, referred to as incremental balancing strategy (IBS), to obtain an optimal load distribution. Based on the rigorous mathematical analysis, a number of interesting and useful properties exhibited by the algorithm are proven. Numerical examples are presented for the ease of understanding.

Wavelength-division multiplexed (WDM) networks using wavelength routing are emerging to be the right choice for the future transport networks. In a WDM-based transport network, the optical layer provides circuit-switched lightpath services to the client layer such as IP, SONET, and ATM. The set of lightpaths in the optical layer defines the virtual topology. Since the optical switches (cross-connects) are reconfigurable, the virtual topology can be reconfigured in accordance with the changing traffic demand pattern at the client layer in order to optimize the network performance. On the other hand, changing the virtual topology can be disruptive to the network since the traffic at each node must be buffered or re-routed while the topology is being reconfigured. We develop a reconfiguration algorithm to reduce the cost of virtual topology reconfiguration in WDM optical ring networks. The algorithm is based on the concept of splitting and merging existing lightpaths, together with cost-benefit analysis to reduce the network reconfiguration cost. Our objective is to reduce the number of lightpaths that need to be reconfigured, while ensuring that the network congestion is low. The performance of the algorithm is verified through simulation experiments.

In this paper, the design and development of a web based text-to-sign language translator, Sign Animator, is discussed. To represent signs accurately, the system uses rotational angles to specify arm location and hand shapes, and uses interpolation between two arm locations to generate the movements within the signs. Some form of collision detection and limits on rotational angles is used to prevent awkward positioning of the humanoid. This paper will present a feasibility study and describe the development of the prototype text-to-sign translation system.

Optimal distribution of divisible loads in bus networks is considered in this paper. The problem of minimizing the processing time is investigated by including all the overhead components that could penalize the performance of the system, in addition to the inherent communication and computation delays. These overheads are considered to be constant additive factors to the respective communication and computation components. Closed-form solution for the processing time is derived and the influence of overheads on the optimal processing time is analyzed. We derive a necessary and sufficient condition for the existence of the optimal processing time. We then study the effect of changing the load distribution sequence on the time performance. Through rigorous analysis, an optimal sequence to distribute the load among the processors is identified, whenever it exists. In case such an optimal sequence fails to exist, we present a greedy algorithm to obtain a suboptimal sequence based on some important properties of the overhead factors. Then, the effect of granularity of the data that is divisible is considered in the analysis for the case of homogeneous networks. An integer approximation algorithm capable of generating integer values of the load fractions in time O(m), where m is the number of processors in the network, is proposed. We then show that the upper bound on the suboptimal solution generated by our algorithm lies within a radius given by the sum of the computation and communication delays. Several numerical examples are presented to illustrate the concepts.

The problem of optimal divisible load distribution in distributed bus networks employing a heterogeneous cluster of processors is addressed. The objective is to minimize the total processing time of the entire load subject to the communication and computation delays. In the mathematical model we adopt, both the granularity of the load fractions and all the associated overheads (also referred to as start-up costs) in the process of communication and computation, are considered explicitly in the problem formulation. We introduce a directed flow graph model for representing the load distribution process. This representation is novel to this literature. With this model, we first derive a closed-form solution for an optimal processing time. We propose an integer approximation algorithm and derive ultimate performance bounds for the class of homogeneous networks. We then extend the problem to a special class of application problems in which the data partitioning is restricted to a finite number of partitions. For this case, we present a recursive procedure to obtain optimal processing time. We then present two different integer approximation algorithms-PIA and IIA that could generate integer load fractions and yield suboptimal solutions. The choice of these algorithms are also analyzed. All the results are extended to a class of homogeneous networks to obtain ultimate performance bounds. Several illustrative examples are provided for ease of explanation.

The problem of obtaining optimal processing time in a distributed computing system consisting of (N+1) processors and N communication links, arranged in a single-level tree architecture, is considered. It is shown that optimality can be achieved through a hierarchy of steps involving optimal load distribution, load sequencing, and processor-link arrangement. Closed-form expressions for optimal processing time is derived for a general case of networks with different processor speeds and different communication link speeds. Using these closed-form expressions, the paper analytically proves a number of significant results that in earlier studies were only conjectured from computational results. In addition, it also extends these results to a more general framework. The above analysis is carried out for the cases in which the root processor may or may not be equipped with a front-end processor. Illustrative examples are given for all cases considered.>