Explore the vast range of titles available.

Cloud computing enables the execution of various applications submitted by the users in the virtualized Cloud environment. However, the Cloud infrastructure consumes a significant amount of electrical energy to provide services to its users that have a detrimental effect on the environment. Many of these applications (tasks), like those belonging to the healthcare system, scientific research, the Internet of Things (IoT), and others, are deadline-sensitive. Hence efficient scheduling of tasks is essential to prevent deadline violation, decrease makespan, and at the same time reduce energy consumption. To address this issue, we have considered the bi-objective optimization problem of minimization of energy and makespan and have proposed two scheduling approaches for independent, deadline-sensitive tasks in a heterogeneous Cloud environment. Our first approach is a greedy heuristic based on the Linear Weighted Sum technique. The second one is based on Ant Colony Optimization and uses a combination of heuristic search and positive feedback of information to improve the solution. Both approaches use a three-tier model where tasks are scheduled by taking into account the properties of three entities- tasks, VMs, and hosts. Moreover, we have proposed a suitable strategy for scaling of Cloud resources to improve energy-efficiency and task schedulability. Extensive simulations using Google Cloud trace-logs and comparison with some state-of-art approaches validate the effectiveness of our proposed scheduling techniques in achieving a proper trade-off between the energy consumption of the virtualized Cloud infrastructure and the average makespan of the tasks.

As the number of enterprises dispatching their workload to the cloud has increased significantly over the last decade, service level agreements (SLAs) becoming a key element to consider for maintaining the quality of service (QoS). In order to facilitate the perseverance of service quality at a satisfactory level, clouds perform load balancing through migration of virtual machines (VMs) from overloaded physical machines (PMs). However, there are several challenges in achieving effective and efficient load balancing. First, VMs in clouds use different resources to serve a variety of applications, which results in varying levels of resource overutilization in different PMs. Second, due to the application’s time-varying heterogeneous nature of resource requirements, the PM’s resource consumption vary over time, making the profiling of resources difficult. Migration decisions in previous load balancing techniques are mostly based on deterministic resource demand estimation, which treats each resource equally and leads to inefficient migrations, causing severe SLA violations in terms of performance degradation. To address this problem, we propose a SLA-aware stochastic load balancing scheme using VM migrations, namely SLA-LB. It provides probabilistic guarantee against resource overloading, while satisfying the SLA. As opposed to previous methods, SLA-LB dynamically assigns different weights to different resources based on PM’s overload probability and effectively addresses the multidimensional resource requirement with stochastic characterization. Experimental results using PlanetLab and Google Cluster trace show that SLA-LB outperforms previous load balancing methods, i.e., RIAL, Sandpiper and CloudScale in terms of performance degradation by an average margin of 10.8%, 23.53% and 33%, respectively.

Convergecast is the process of gathering all sensor data at the base station. Convergecast in a Sensor-Cloud is achieved by creating multiple data-gathering trees over collaborating wireless sensor networks and collecting sensor data via those trees. Designing a schedule for convergecast can be one-shot (only for one round of data collection) or pipelined (data collected at regular intervals repeatedly). We give a pipelined algorithm for convergecast, assuming a raw data model (no aggregation). Firstly, we design a slot allocation algorithm considering only the adjacency constraints and traversing the trees in BFS (breadth first search) order. Our schedule requires at most M+H+1 time slots for the first round of data collection and at most M+1 slots in subsequent rounds, where M is the lower bound to complete one round of convergecast and H is the maximum height of the data gathering trees. Our schedule does not require any node to store more than two packets. In the next step, we consider interference among simultaneous transmissions and allocate multiple frequency channels to ensure interference below a certain threshold. The problem of minimizing the frequency channels while maintaining SINR above the desired value is reduced to a semidefinite programming relaxation. In our simulation, the time for the first round of convergecast is much lower than M+H+1 (closer to M+1), and the channel allocation algorithm requires either optimal or just one more than the optimal number of channels.

Multiple object detection and tracking play a very crucial role in solving several elementary problems in real-time surveillance video analysis and computer vision. However, it is a challenging problem because real-time surveillance videos are typically affected by a variety of adverse environmental effects. In this work, we propose a novel surveillance framework, called a unifocal motion tracking surveillance system (UMTSS), for multi-object tracking in real-time videos. The proposed UMTSS combines two significant steps. First, a Faster-RCNN with inception-v2 model is employed here to detect multi-objects efficiently in each video frame. Then, a unifocal feature-based KLT (Kanade-Lucas-Tomasi) method is proposed for tracking objects across the video frames based on region proposals generated by the object detector in the previous phase. Also, we have proposed a new tracking parameter, called dynamic tracking accuracy (DTA), to quantify the performance of the tracking algorithms. The performance of our UMTSS has been evaluated on five standard crowd video databases, namely CrowdHuman, PETS, UCSD, AGORASET and CRCV, and compared with state-of-the-art methods in terms of different qualitative and quantitative measures. It has been observed that our UMTSS outperforms the state-of-the-art methods.

An accurate host power model is necessary for effective power management in data centers which is crucial for reducing energy consumption and cost. One should evaluate the power models for different workloads and host configurations. We have analyzed several existing power models by varying the workload type (CPU, memory, and disk-intensive) and host configurations. By analyzing the system performance and nature of the power consumption of the hosts, we have identified some performance counter parameters that determine the system power consumption. We have proposed three power models based on multi-variable Linear Regression, Support Vector Regression (SVR), and Artificial Neural Network (ANN). Experimental results show that compared to the existing models, our proposed power models, especially those based on SVR and ANN, more accurately predict the power consumption of the hosts. We have also conducted simulation experiments to show the importance of the power model in the energy-efficient resource management of the hosts in the data center. Results show that the use of our SVR-based and ANN-based power models in a resource management approach can effectively decrease the energy consumption of the data center. Moreover, we have proposed an energy-efficient virtual machine (VM) placement and consolidation algorithm that further reduces energy consumption. At first, we formulated a model using integer linear programming. Then, we designed a heuristic based on Vogel’s Approximation Method. Extensive simulation on the CloudSim platform with benchmark workload data and the Google Cloud trace logs shows that our approach outperforms the state-of-the-art algorithms under comparison in terms of energy efficiency and quality of service (QoS). The results also highlight the importance of a suitable VM placement and consolidation approach and an accurate power model in reducing energy consumption.

Workflows are special applications used to solve complex scientific problems. The emerging Workflow as a Service (WaaS) model provides scientists with an effective way of deploying their workflow applications in Cloud environments. The WaaS model can execute multiple workflows in a multi-tenant Cloud environment. Scheduling the tasks of the workflows in the WaaS model has several challenges. The scheduling approach must properly utilize the underlying Cloud resources and satisfy the users’ Quality of Service (QoS) requirements for all the workflows. In this work, we have proposed a heurisine-sensitive workflows in a containerized Cloud environment for the WaaS model. We formulated the problem of minimizing the MIPS (million instructions per second) requirement of tasks while satisfying the deadline of the workflows as a non-linear optimization problem and applied the Lagranges multiplier method to solve it. It allows us to configure/scale the containers’ resources and reduce costs. We also ensure maximum utilization of VM’s resources while allocating containers to VMs. Furthermore, we have proposed an approach to effectively scale containers and VMs to improve the schedulability of the workflows at runtime to deal with the dynamic arrival of the workflows. Extensive experiments and comparisons with other state-of-the-art works show that the proposed approach can significantly improve resource utilization, prevent deadline violation, and reduce the cost of renting Cloud resources for the WaaS model.

A crucial research problem in the field of wireless sensor network is to maximize its lifetime. One approach to solve this problem is to group the nodes in clusters or chains. One of the nodes in a cluster or chain takes the responsibility of collecting data from the other sensors in its group and sends the data to the base station or sink. As the energy spent per round is comparatively higher for a cluster head(CH)/chain leader (CL), the lifetime of the network can increase if we rotate the role of CH/CL among the nodes in the cluster/chain. In this paper, we have proposed several algorithms to select CH/CL at different rounds to maximize the lifetime of the WSN. We have shown that maximizing lifetime measured as first node die, can be solved optimally using an integer linear program. Then we have given a polynomial-time algorithm to solve this problem by relaxing to a linear programming problem. We have also given algorithms to maximize the lifetimes measured as half of the nodes die and last node dies. Simulation results indicate significant improvement in the performance of our proposed algorithms compared to BP-DCA, NEECP and naive Bayesian-based algorithm.

Mobile crowd sensing (MCS) is an emerging sensing platform that concedes mobile users to efficiently collect data and share information with the MCS service providers. Despite its benefits, a key challenge in MCS is how beneficially select a minimum subset of participants from the large user pool to achieve the desired level of coverage. In this paper, we propose several algorithms to choose a minimum number of mobile users(or participants) who met the desired level of coverage. We consider two different cases, in the first case, only a single participant is allowed to upload a data packet for a particular target, whereas for the other case, two participants are allowed to do the same (provided that the target is covered by more than one participants). An optimal solution to the problem can be found by solving integer linear programmings (ILP’s). However, due to the exponential complexity of the ILP problem, for the large input size, it is infeasible from the point of execution time as well as the requirement of having the necessary information about all the participants in a central location. We also propose a distributed participant selection algorithm considering both the cases, which are dynamic in nature and run at every user. Each user exchanges their message with the neighbors to decide whether to remain idle or active. A series of experiments are executed to measure the performance of the proposed algorithms. Simulation results reveal the proximity of the proposed distributed algorithm compared to the optimal result providing the same coverage.

Efficient transmission of prioritized data packets is of utmost importance for wireless body area networks (WBAN). In this paper, the problem of inter-WBAN communication with multiple edges is investigated to maximize the transmission efficiency subject to data priority and channel quality. The contribution is twofold. First, the authors jointly consider the channel quality, receiver type, and the remaining energy of the forwarders to frame a fuzzy analytical hierarchy process for the prioritization of the suitable forwarder. A forwarder can be a relay node or an edge where the WBAN coordinators send their data for further processing. Second, a transmission power selection mechanism is also proposed subject to data priority and channel quality through employing a fuzzy inference system. Numerical results reveal that the proposed model is able to dynamically select the suitable node (relay or edge) for data communication subject to varying channel conditions while guaranteeing minimal transmission power for successful data delivery.

Background High-throughput Next-Generation Sequencing (NGS) techniques are advancing genomics and molecular biology research. This technology generates substantially large data which puts up a major challenge to the scientists for an efficient, cost and time effective solution to analyse such data. Further, for the different types of NGS data, there are certain common challenging steps involved in analysing those data. Spliced alignment is one such fundamental step in NGS data analysis which is extremely computational intensive as well as time consuming. There exists serious problem even with the most widely used spliced alignment tools. TopHat is one such widely used spliced alignment tools which although supports multithreading, does not efficiently utilize computational resources in terms of CPU utilization and memory. Here we have introduced PVT (Pipelined Version of TopHat) where we take up a modular approach by breaking TopHat’s serial execution into a pipeline of multiple stages, thereby increasing the degree of parallelization and computational resource utilization. Thus we address the discrepancies in TopHat so as to analyze large NGS data efficiently. Results We analysed the SRA dataset (SRX026839 and SRX026838) consisting of single end reads and SRA data SRR1027730 consisting of paired-end reads. We used TopHat v2.0.8 to analyse these datasets and noted the CPU usage, memory footprint and execution time during spliced alignment. With this basic information, we designed PVT, a pipelined version of TopHat that removes the redundant computational steps during ‘spliced alignment’ and breaks the job into a pipeline of multiple stages (each comprising of different step(s)) to improve its resource utilization, thus reducing the execution time. Conclusions PVT provides an improvement over TopHat for spliced alignment of NGS data analysis. PVT thus resulted in the reduction of the execution time to ~23% for the single end read dataset. Further, PVT designed for paired end reads showed an improved performance of ~41% over TopHat (for the chosen data) with respect to execution time. Moreover we propose PVT-Cloud which implements PVT pipeline in cloud computing system.