Search

Motivated by applications in data center networks, in this paper, we study the problem of scheduling in an input queued switch. While throughput maximizing algorithms in a switch are well-understood, delay analysis was developed only recently. It was recently shown that the well-known MaxWeight algorithm achieves optimal scaling of mean queue lengths in steady state in the heavy-traffic regime, and is within a factor less than 2 of a universal lower bound. However, MaxWeight is not used in practice because of its high time complexity. In this paper, we study several low complexity algorithms and show that their heavy-traffic performance is identical to that of MaxWeight. We first present a negative result that picking a random schedule does not have optimal heavy-traffic scaling of queue lengths even under uniform traffic. We then show that if one picks the best among two matchings or modifies a random matching even a little, using the so-called flip operation, it leads to MaxWeight like heavy-traffic performance under uniform traffic. We then focus on the case of non-uniform traffic and show that a large class of low time complexity algorithms have the same heavy-traffic performance as MaxWeight, as long as it is ensured that a MaxWeight matching is picked often enough. We also briefly discuss the performance of these algorithms in the large scale heavy-traffic regime when the size of the switch increases simultaneously with the load. Finally, we perform empirical study on a new algorithm to compare its performance with some existing algorithms. [ABSTRACT FROM AUTHOR]

This paper presents the first network configuration verifier that provides fast all-pair reachability analysis of incremental configuration changes for network overlay data center networks (DCNs). Network overlay DCNs leverage distributed routing protocol on edge leaf switches to disseminate overlay routes and establish overlay tunnels. In addition, network overlay DCNs use access control lists, microsegmentation policy, policy-based routing and firewall policy to control east-west and north-south traffic. Although some incremental verification approaches have been proposed, they either do not support certain forwarding features of the network, or are not efficient. Our configuration verifier addresses these issues through the following components: 1) a port predicate based forwarding model that is general to support all features; 2) fine-grained association technique to index possibly affected reachable pairs by changed interfaces in the original network; and 3) required waypoint path computation that finds all reachable pairs related to changed interfaces in the new network. Based on these components, our verifier presents two incremental verification algorithms that are specially designed for different service update cases. Experiment results show that our incremental verification algorithms are accurate and fast. For all-pair reachability, our verifier performs change-impact analysis within 15s for networks with 200 leafs (4000 subnets and 16 million pairs), outperforming existing approaches by up to 10x. [ABSTRACT FROM AUTHOR]

In cloud computing (CC), load balancing tasks remain a critical problem in distributing resources from a data center. Ensure that every virtual machine (VM) has a balanced load to maximize capacity utilization. In the CC world, load balancing is a Non-Polynomial (NP) problem resolved with metaheuristic algorithms. A new Quasi-Oppositional Dragonfly Algorithm for Load Balancing (QODA-LB) has been developed to obtain optimum resource scheduling in a CC configuration. The proposed QODA-LB algorithm uses three variables to calculate an objective function: execution time, execution cost, and charge. The QODA-LB algorithm assigns tasks to VM according to its potential and the resulting objective function. Also, the QODA-LB algorithm employs the Quasi-Oppositional Based Learning principle to increase the standard convergence rate of the Dragonfly (DA) algorithm. A complete series of experiments were conducted, and the results were analyzed in various ways to ensure the increased efficiency of the QODA-LB algorithm. Simulation results demonstrated an optimal load balancing efficiency and outperformed key approaches. [ABSTRACT FROM AUTHOR]

Cloud computing provides unprecedented advantages of using computing resources with very less efforts and cost. The energy utilization in cloud data centers has forced the cloud service providers to raise the expense of using its services and has increased the carbon footprints in the environment. Many static bin-packing algorithms exist which can reduce energy by some percentage, but with new era of digitization, advanced and dynamic techniques are required which can serve heterogeneous users and random users' requests. Thus, in this paper, two new dynamic best-fit decreasing-based bin-packing algorithms are proposed wherein the first technique is for service providers and focuses on increasing server utilization and the second approach acts as a switcher to harness best results among all algorithms. Both techniques deliberately achieve high performance in terms of total energy consumption, resource utilization, and makespan along with serving continuous and varying requests from customers. The simulations are performed using Java. The results exhibited that DEE-BFD can escalate resource utilization by 96% and EM switcher can reduce total energy consumption by 49% and reduce makespan by 56%. [ABSTRACT FROM AUTHOR]

Copyright of Telecommunication Engineering is the property of Telecommunication Engineering and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Velocity dealiasing is an essential task for correcting the radial velocity data collected by Doppler radar. To improve the accuracy of velocity dealiasing, traditional dealiasing algorithms usually set a series of empirical thresholds, combine three- or four-dimensional data, or introduce other observation data as a reference. In this study, we transform the velocity dealiasing problem into a clustering problem and solve this problem using the density-based spatial clustering of applications with noise (DBSCAN) method. This algorithm is verified with a case study involving radar data on the tropical cyclone Mangkhut in 2018. The results show that the accuracy of the proposed algorithm is close to that of the four-dimensional dealiasing (4DD) method proposed by James and Houze; yet, it only requires two-dimensional velocity data and eliminates the need for other reference data. The results of the case study also show that the 4DD algorithm filters out many observation gates close to the missing data or radar center, whereas the proposed algorithm tends to retain and correct these gates. [ABSTRACT FROM AUTHOR]

• Right-sizing of heterogeneous data centers with d different server types. • Discrete setting, time- and load-independent operating cost. • 2d-competitive deterministic online algorithm. • 1.582d-competitive randomized online algorithm. • Lower bound of 2d for deterministic online algorithms. Power consumption is the major cost factor in data centers. It can be reduced by dynamically right-sizing the data center according to the currently arriving jobs. If there is a long period with low load, servers can be powered down to save energy. For identical machines, the problem has already been solved optimally by Lin et al. (2013) and Albers and Quedenfeld (2018). In this paper, we study how a data-center with heterogeneous servers can dynamically be right-sized to minimize the energy consumption. There are d different server types with various operating and switching costs. We present a deterministic online algorithm that achieves a competitive ratio of 2 d as well as a randomized version that is 1.58 d -competitive. Furthermore, we show that there is no deterministic online algorithm that attains a competitive ratio smaller than 2 d. Hence our deterministic algorithm is optimal. In contrast to related problems like convex body chasing and convex function chasing, we investigate the discrete setting where the number of active servers must be integral, so we gain truly feasible solutions. [ABSTRACT FROM AUTHOR]

Recently, the virtual machine deployment algorithm uses physical machine less or consumes higher energy in data centers, resulting in declined service quality of cloud data centers or rising operational costs, which leads to a decrease in cloud service provider's earnings finally. According to this situation, a resource clustering algorithm for cloud data centers is proposed. This algorithm systematically analyzes the cloud data center model and physical machine's use ratio, establishes the dynamic resource clustering rules through k-means clustering algorithm, and deploys the virtual machines based on clustering results, so as to promote the use ratio of physical machine and bring down energy consumption in cloud data centers. The experimental results indicate that, regarding the compute-intensive virtual machines in cloud data centers, compared to contrast algorithm, the physical machine's use ratio of this algorithm is improved by 12% on average, and its energy consumption in cloud data center is lowered by 15% on average. Regarding the general-purpose virtual machines in cloud data center, compared to contrast algorithm, the physical machine's use ratio is improved by 14% on average, and its energy consumption in cloud data centers is lowered by 12% on average. Above results demonstrate that this method shows a good effect in the resource management of cloud data centers, which may provide reference to some extent. [ABSTRACT FROM AUTHOR]

In the past decade, the cloud environment has emerged as a resource-enriched technology for computational modeling. Most cloud data centers are deployed far from the device layer, which may cause higher service delivery time. Therefore, Fog servers are implemented to retain the commitment of minimum service delivery time. The computational resources (processing unit, memory, storage) and network bandwidth are to be appropriately scheduled to make an efficient system. For delay and resource-constrained devices, the essential issues include minimum service latency and improved system performance by offloading non-sensitive devices to the cloud layer. In this study, an adaptive bandwidth adjustment technique is proposed considering the sensitivity of the applications and load on the multiple queues at the access point based on which, a compelling dynamic bandwidth ratio is calculated. Further, blocking probability of task and task dropping probability is used as task offloading probability to overcome the dropping ratio of type-II applications. The random early detection (RED) algorithm is implemented in queueing management to maximize the system performance. Extensive simulation results highlighted that the proposed model significantly improves the system performance in terms of network delay, queue delay, and overall response time over traditional algorithms. [ABSTRACT FROM AUTHOR]

A huge cloud data center makes it possible to offer computing as a utility to customers. However, the main challenge is to fulfill the customer's dynamic workload requirement seamlessly. Additionally, cloud data center consumes an enormous amount of power due to improper scheduling of virtual machines over the physical machines, which lead to inefficient usage of heterogeneous computing resources. So, to minimize energy consumption in the cloud data center, virtual machines should be scheduled in an energy-efficient way. In this paper, an Artificial Immune System based Virtual Machine Scheduling using Modified Clonal Selection Algorithm (VMS-MCSA) is proposed to schedule virtual machines energy efficiently. The classical Clonal Selection Algorithm(CSA) operators are modified such that they can be applied to the discrete optimization dynamic virtual machine scheduling problem. The randomized mutation operator is proposed, which reschedule VMs at each scheduling interval to handle the dynamicity of workload with minimum virtual machine migrations. Additionally, the VM-consolidation model was proposed for constraint-based virtual machine migration. The proposed VMS-MCSA algorithm is implemented on a cloudsim simulator, and the results show that the VM scheduling using VMS-MCSA is energy-efficient compared to other recent approaches. [ABSTRACT FROM AUTHOR]

With the increment in figuring advancements, Cloud computing has permitted the clients to get to assets from anyplace whenever and to pay for the assets based on pay-peruse. Cloud computing allocates the submitted tasks to virtual machine (VM) by dynamic allocation. With the bandwidth and network connectivity getting better, more number of users and organizations are moving to Cloud. All things considered, there are different difficulties that are looked by Cloud service provider. One of the most important challenges is of load balancing. With the number of users gradually increasing, load balancing has become more important. Numerous calculations have been proposed till now for load adjusting, where every one of it centres around the various boundaries. However, these techniques suffer from various issues like - stuck in native optima. None of the calculations has demonstrated to be totally proficient. To deal with the issues identified with existing meta-heuristic procedures, in this paper, a mutation based ant colony optimization load balancing algorithm is proposed. Here we focus on the mutation in Ant Colony Optimization (ACO) algorithm for efficiently balancing the load among various data centers to further develop the presentation boundaries like response time and while improving the overall fitness function. [ABSTRACT FROM AUTHOR]

With the development of big data technologies, green cloud data centers have become a key factor in academia and industry. An energy-efficient cloud data center can save costs for cloud computing users. However, the problem of virtual machine mapping has always been a core problem. In the most existing research, the energy consumption generated by cloud data centers has become an important bottleneck restricting the technology of cloud computing. This paper establishes a consumption model of cloud data center energy and a virtual machine mapping rule. Based on this, a multi-dimensional double descending maximum padding priority (MD3MP2) virtual machine mapping algorithm is proposed. The algorithm can not only solve the one-dimensional virtual machine mapping problem of homogeneous data centers, but also successfully solve the multi-dimensional virtual machine mapping problem of homogeneous data centers. Finally, the algorithm is compared with four other algorithms. The experimental results show that the MD3MP2 algorithm is better than the compared algorithms. [ABSTRACT FROM AUTHOR]

Cloud computing applications have grown rapidly in the last decade, where many organizations are migrating their applications to cloud data center as they expected high performance, reliability, and the best quality of service. Data centers deploy a variety of technologies, such as software-defined networks (SDN), to effectively manage their resources. The SDN approach is a highly flexible network architecture that automates network configuration using a centralized controller to overcome traditional network limitations. This paper proposes an SDN-based monitoring algorithm to detect the performance anomaly and identify the bottleneck of the distributed application in the cloud data center using the support vector machine algorithm. It collects the data from the network devices and calculates the performance metrics for the distributed application components that are used to train the SVM algorithm and build a baseline model of the normal behavior of the distributed application. The SVM model detects performance anomaly behavior and identifies the root cause of bottlenecks using one-class support vector machine (OCSVM) and multi-class support vector machine (MCSVM) algorithms. The proposed method does not require any knowledge about the running applications or depends on static threshold values for performance measurements. Simulation results show that the proposed method can detect and locate the failure occurrences efficiently with high precision and low overhead compared to statistical methods, Naive Bayes Classifier and the decision tree machine learning method. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Harbin University of Science & Technology is the property of Journal of Harbin University of Science & Technology and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

In this study, a re-tuned algorithm based on the geophysical model function (GMF) C-SARMOD2 is proposed to retrieve wind speed from Synthetic Aperture Radar (SAR) imagery collected by the Chinese C-band Gaofen-3 (GF-3) SAR. More than 10,000 Vertical-Vertical (VV) and Horizontal-Horizontal (HH) polarization GF-3 images acquired in quad-polarization stripmap (QPS) and wave (WV) modes have been collected during the last three years, in which wind patterns are observed over open seas with incidence angles ranging from 18° to 52°. These images, collocated with wind vectors from the European Centre for Medium-Range Weather Forecast (ECMWF) reanalysis at 0.125° resolution, are used to re-tune the C-SARMOD2 algorithm to specialize it for the GF-3 SAR (CSARMOD-GF). In particular, the CSARMOD-GF performs differently from the C-SARMOD2 at low-to-moderate incidence angles smaller than about 34°. Comparisons with wind speed data from the Advanced Scatterometer (ASCAT), Chinese Haiyang-2B (HY-2B) and buoys from the National Data Buoy Center (NDBC) show that the root-mean-square error (RMSE) of the retrieved wind speed is approximately 1.8 m/s. Additionally, the CSARMOD-GF algorithm outperforms three state-of-the-art methods – C-SARMOD, C-SARMOD2, and CMOD7 – that, when applied to GF-3 SAR imagery, generating a RMSE of approximately 2.0–2.4 m/s. [ABSTRACT FROM AUTHOR]

The proliferation of Internet-of-Things (IoT) devices and cloud-computing applications over siloed data centers is motivating renewed interest in the collaborative training of a shared model by multiple individual clients via federated learning (FL). To improve the communication efficiency of FL implementations in wireless systems, recent works have proposed compression and dimension reduction mechanisms, along with digital and analog transmission schemes that account for channel noise, fading, and interference. The prior art has mainly focused on star topologies consisting of distributed clients and a central server. In contrast, this paper studies FL over wireless device-to-device (D2D) networks by providing theoretical insights into the performance of digital and analog implementations of decentralized stochastic gradient descent (DSGD). First, we introduce generic digital and analog wireless implementations of communication-efficient DSGD algorithms, leveraging random linear coding (RLC) for compression and over-the-air computation (AirComp) for simultaneous analog transmissions. Next, under the assumptions of convexity and connectivity, we provide convergence bounds for both implementations. The results demonstrate the dependence of the optimality gap on the connectivity and on the signal-to-noise ratio (SNR) levels in the network. The analysis is corroborated by experiments on an image-classification task. [ABSTRACT FROM AUTHOR]

The interest in selecting an appropriate cloud data center is exponentially increasing due to the popularity and continuous growth of the cloud computing sector. Cloud data center selection challenges are compounded by ever-increasing users' requests and the number of data centers required to execute these requests. Cloud service broker policy defines cloud data center's selection, which is a case of an NP-hard problem that needs a precise solution for an efficient and superior solution. Differential evolution algorithm is a metaheuristic algorithm characterized by its speed and robustness, and it is well suited for selecting an appropriate cloud data center. This paper presents a modified differential evolution algorithm-based cloud service broker policy for the most appropriate data center selection in the cloud computing environment. The differential evolution algorithm is modified using the proposed new mutation technique ensuring enhanced performance and providing an appropriate selection of data centers. The proposed policy's superiority in selecting the most suitable data center is evaluated using the CloudAnalyst simulator. The results are compared with the state-of-arts cloud service broker policies. [ABSTRACT FROM AUTHOR]

Management of IT services is rapidly adapting to the cloud computing environment due to optimized service delivery models. Geo distributed cloud data centers act as a backbone for providing fundamental infrastructure for cloud services delivery. Conversely, their high growing energy consumption rate is the major problem to be addressed. Cloud providers are in a hunger to identify different solutions to tackle energy management and carbon emission. In this work, a multi-cloud environment is modeled as geographically distributed data centers with varying solar power generation corresponding to its location, electricity price, carbon emission, and carbon tax. The energy management of the workload allocation algorithm is strongly dependent on the nature of the application considered. The task deadline and brownout information is used to bring in variation in task types. The renewable energy-aware workload allocation algorithm adaptive to task nature is proposed with migration policy to explore its impact on carbon emission, total energy cost, brown and renewable power consumption. [ABSTRACT FROM AUTHOR]

The advent of virtualization technology has created a huge potential application for cloud computing. In virtualization, a large hardware resource is often broken down into smaller virtual units. These small units are then provisioned to different clients. However, these services need to be provided in such a way that resources are properly utilized. To achieve this, many of the scheduling, allocation, and provisioning issues of data centers are formulated as optimization problems. The virtual machine placement problem (VMPP) is a typical provisioning problem of data centers. In VMPP, several virtual machine requests are to be hosted on physical machines such that a minimum number of physical machines are used. This work proposes a cuckoo search (CS) inspired algorithm for solving the VMPP. To improve the algorithm's performance, new cost and perturbation functions are developed. The proposed method was tested on two well-known benchmark datasets. It outperformed the reordered grouping genetic algorithm, best-fit decreasing, first-fit decreasing, and an earlier CS method called multiCSA. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Jilin University (Science Edition) / Jilin Daxue Xuebao (Lixue Ban) is the property of Zhongguo Xue shu qi Kan (Guang Pan Ban) Dian zi Za zhi She and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Spatial prediction is commonly achieved under the assumption of a Gaussian random field by obtaining maximum likelihood estimates of parameters, and then using the kriging equations to arrive at predicted values. For massive datasets, fixed rank kriging using the expectation–maximization algorithm for estimation has been proposed as an alternative to the usual but computationally prohibitive kriging method. The method reduces computation cost of estimation by redefining the spatial process as a linear combination of basis functions and spatial random effects. A disadvantage of this method is that it imposes constraints on the relationship between the observed locations and the knots. We develop an alternative method that utilizes the spatial mixed effects model, but allows for additional flexibility by estimating the range of the spatial dependence between the observations and the knots via an alternating expectation conditional maximization algorithm. Experiments show that our methodology improves estimation without sacrificing prediction accuracy while also minimizing the additional computational burden of extra parameter estimation. The methodology is applied to a temperature dataset archived by the United States National Climate Data Center, with improved results over previous methodology. [ABSTRACT FROM AUTHOR]

Summary: Data center networks (DCNs) are facing challenging control problems such as flow scheduling, congestion control, load balancing, and bandwidth allocation when dynamically handling heterogeneous mice and elephant flows. This paper focuses on the load balancing in DCNs as an online decision‐making problem with high complexity (NP‐hard problem). Traditional load balancing approaches are usually the heuristic algorithms relying on the operator's viewpoint and a short‐term knowledge of the traffic conditions and network environment. This paper proposes DeepRLB, a deep reinforcement learning (DRL)‐based load balancing approach for software‐defined networking‐based DCNs which exploits the deep deterministic policy gradient (DDPG) algorithm to adaptively learn the link‐weight values by observing the traffic flow characteristics. The actor and critic neural networks in the DDPG model have been designed based on two learning models: fully connected (DeepRLB‐Full) and convolutional (DeepRLB‐Conv). The output of the model is further used by the controller to determine the forwarding paths for traffic flows accordingly. We evaluated the performance of DeepRLB by comparing it with ECMP as well as a per‐packet load balancing algorithm in three different data center topologies under 30 heterogeneous traffic demand matrices. The obtained results showed that the DeepRLB algorithm outperforms the ECMP with respect to studied load balancing metrics considerably (DeepRLB‐Conv achieved the best performance). As a sample, DeepRLB‐Full and DeepRLB‐Conv were able to reach 19.14% and 24.46% reduction in the total number of over‐utilized and under‐utilized network links compared to ECMP, respectively. [ABSTRACT FROM AUTHOR]

In order to satisfy the rapidly increasing demand for data volume, large data center networks (DCNs) have been proposed. In 2019, Zhang et al. proposed a new highly scalable DCN architecture named HSDC (Highly Scalable Data Center Network), which can achieve greater incremental scalability. In this paper, we give the definition of the logical graph of HSDC, named H n , which can be treated as a compound graph of hypercube and complete graph of the same dimension. First, we prove that the connectivity and tightly super connectivity of H n are both n. Then, we give an O(n) shortest unicast algorithm to find a shortest path between any two distinct nodes in H n , and prove the correctness of this algorithm. In fact, we also prove that the length of the shortest path constructed by this algorithm is no more than 2 d + 1 if d < n and at most 2d if d = n , where d is the Hamming distance between the start and end nodes, and the diameter of H n is 2n. [ABSTRACT FROM AUTHOR]

Fog computing as a complementary paradigm to cloud computing is a heuristic shift in service delivery that promises a leap in efficiency and flexibility for cloud-based Internet of Things applications. The performance characteristics of cloud/fog computing attract significant attention from researchers lately. One of the critical challenges in this field is controlling and reducing the massive amount of energy consumption in the cloudlets while still maintaining the Service Level Agreement's performance requirements. Many virtual machine (VM) allocation and consolidation strategies are investigated to address the challenges mentioned earlier. However, many of the solutions save energy at the cost of performance degradation. This paper proposes a novel multi-step VM allocation algorithm called enhanced performance-to-power ratio for workflow applications "E-PRWA" in cloud/fog environment. The proposed heuristic algorithm strives to achieve a trade-off between node performance and power consumption. Operating machine hosts at the highest performance-to-power ratio can save a tremendous amount of energy without degrading system performance. The proposed model consists of four stages: (a) detecting overutilized or underutilized nodes based on the preferred utilization (PU); (b) VM selection for migration from the overutilized nodes to underutilized nodes; (c) switching off selected underutilized nodes; (d) deploying the migration VMs based on the modified best-fit decreasing algorithm with PPR, latency overhead, and computational cost consideration. Extensive simulation results illustrate that compared with three baseline energy-efficient VM allocation and selection algorithms, E-PRWA can achieve an average of up to 65.41% of energy-saving with fewer migration number in fog computing. [ABSTRACT FROM AUTHOR]

Unexpected faults result in unscheduled cloud outage, which negatively affects the completion of workflow tasks in the cloud. This paper presents a novel PageRank-based fault handling strategy to rescue workflow tasks at the faulty data center. The proposed approach uses a holistic view and considers the task attributes, the timeline scenario, and the overall cloud performance. A priority assignment system is developed based on the modified PageRank algorithm to prioritise workflow tasks. A min-max normalization method is applied to select the target data center and match the timeline at this data center. Additionally, a dynamic PageRank-constrained task scheduling algorithm is proposed to generate the task scheduling solution. The simulation results show that the proposed approach can achieve better fault handling performance, measured by task resilience ratio, workflow resilience ratio, and workflow continuity ratio in both the traditional 3-replica and the image backup cloud environment. [ABSTRACT FROM AUTHOR]

The growing demand for cloud computing adoption presents more challenges for researchers to make cloud computing more efficient and affordable for infrastructure providers and end users. The management of cloud computing involves investing in IT infrastructure in the first phase and investing in energy, maintenance and space costs in the second phase. However, energy costs account for a large portion of cloud management costs, and saving energy consumption can significantly reduce overall cloud management costs. Server consolidation is a strategy to improve data center energy efficiency and resource utilization. Virtual machine (VM) placement is considered one of the main problems with VM consolidation. The VM placement problem aims to reduce the number of active physical machines in data center to reduce data center power consumption and maintenance costs. However, the waste of data center resources has a significant impact on the energy efficiency of the data center, so it should be considered in the VM placement strategy. This paper proposes a new method based on the Monarch Butterfly Optimization algorithm (MBO) called MBO-VM for new virtual machine placement, designed to maximize packaging efficiency and reduce the number of active physical servers. CloudSim toolkit is used to test the efficiency of the proposed MBO-VM approach under real cloud workloads as well as synthetic workloads. Simulation results show that MBO-VM produces significantly better results compared with known VM placement techniques. The proposed MBO-VM can reduce the number of active servers more effectively and maximize the packaging efficiency. [ABSTRACT FROM AUTHOR]

Nowadays, cloud computing is known as an internet-based modern area among emerging technologies that brings up an environment, in which computing resources such as hardware, software, storage, etc. can be rented by cloud users based on a pay per use model. Since the size of cloud computing is widely expanding and the number of cloud users is also increasing day by day, high energy consumption becomes a serious concern in the operation of complex cloud data centers. In this regards, Virtual Machine (VM) consolidation plays a vital role in utilizing cloud resources in an efficient manner. It migrates the running VMs from overloaded Physical Machines (PMs) to other PMs considering multiple factors, such as migration overhead, energy consumption, resource utilization, and migration time. Since the VM consolidation issue is known as an NP-hard problem, various nature‐inspired meta-heuristic algorithms aiming to solve this problem have been utilized in recent years. However, a lack of systematic and detailed survey study in this field is obvious. Therefore, this gap motivated us to provide the current paper aiming to highlight the role of nature-inspired meta-heuristic algorithms in the VM consolidation problem, review the existing approaches, offer a detailed comparison of approaches based on important factors, and finally, outline the future directions. [ABSTRACT FROM AUTHOR]

Traditional data centers are shifted toward the cloud computing paradigm. These data centers support the increasing demand for computational and data storage that consumes a massive amount of energy at a huge cost to the cloud service provider and the environment. Considerable energy is wasted to constantly operate idle virtual machines (VMs) on hosts during periods of low load. Dynamic consolidation of VMs from overloaded or underloaded hosts is an effective strategy for improving energy consumption and resource utilization in cloud data centers. The dynamic consolidation of VM from an overloaded host directly influences the service level agreements (SLAs), utilization of resources, and quality of service (QoS) delivered by the system. We proposed an algorithm, namely, GradCent, based on the Stochastic Gradient Descent technique. This algorithm is used to develop an upper CPU utilization threshold for detecting overloaded hosts by using a real CPU workload. Moreover, we proposed a dynamic VM selection algorithm called Minimum Size Utilization (MSU) for selecting the VMs from an overloaded host for VM consolidation. GradCent and MSU maintain the trade-off between energy consumption minimization and QoS maximization under specified SLA goal. We used the CloudSim simulations with real-world workload traces from more than a thousand PlanetLab VMs. The proposed algorithms minimized energy consumption and SLA violation by 23% and 27.5% on average, compared with baseline schemes, respectively. [ABSTRACT FROM AUTHOR]

With the explosive growth in the scale of cloud computing infrastructures, reliability and energy efficiency have become important concerns considering the great complexity of cloud data centers. There is an urgent need for efficient task assignment that can dispatch tasks to appropriate cloud data center servers, which is critical to achieve reliability and energy efficiency in current cloud data centers. Most of the research on task assignment focuses on only one of the objectives of reliability and energy efficiency, while the two objectives are intrinsically conflicting with each other. In this paper, we deal with the problem of task assignment in data centers, with the objective of minimizing the energy consumption while providing failure tolerance to task execution failure. We propose a reliability-aware and energy-efficient task replica assignment algorithm based on running task replicas at a low speed and enabling multiple task replicas to share the same server resources. Each task in a job processed by the cloud computing platform has two instances: main task and task replica (shadow). Each main task runs on an individual server, and the task replica associated with the main task is assigned on a different server. The main tasks run at the full server speed, while the task replicas run at a lower rate than the main tasks. The task replicas can be mapped onto dedicated backup servers or be assigned to the servers on which the main tasks are running. Multiple task replicas can share the same server resources to reduce the number of servers required. We conduct experiments through simulations. Experimental results demonstrate that the proposed algorithm can effectively reduce the energy consumption, while achieving a good balance between the number of servers used and job completion time. [ABSTRACT FROM AUTHOR]

A neural network (NN)‐assisted probabilistic shaping (PS) distribution matcher is proposed, in which the model is simplified by a structured optimization method. The NN algorithm can encode the information sequence, making the signal obey the Gaussian distribution, and can directly restore the received signal. In addition, the algorithm uses the novel training method at both ends of the transmitter and receiver so that the system performance is significantly improved. PS system verification experiments have been carried out under 16QAM‐DMT modulation format. Under the hard decision forward error correction (FEC) threshold of 3.8*10−3 BER, the proposed system achieves 1.1 dB improvement compared to the traditional 16QAM‐DMT system. [ABSTRACT FROM AUTHOR]

Modern virtualized data centers often rely on virtual machine (VM) migrations to consolidate workload on a single machine for energy saving. But VM migrations have many drawbacks, including performance degradation, service disruption etc. Hence, many approaches have been proposed to minimize the overhead when migrations occur. In contrast, this work aims to proactively avoid migrations from happening in the first place. We have proposed a novel consolidation aware scheduling algorithm to minimize the number of migrations for batch processing systems by taking advantage of the prior knowledge of consolidation strategy and job information. We show the problem can be formulated as an integer linear programming (ILP) problem, and an effective heuristic solution can be found by a genetic algorithm. Both real and synthetic workload traces were used to evaluate our methods. Experimental results showed that, after comparing with two popular job scheduling algorithms, our approach has reduced the number of migrations by more than 25%. • Proactively avoid VM migration in virtualized data centers. • A co-design between consolidation and scheduling decisions. • Construct an ILP formulation to find an optimal solution. • Apply genetic algorithm to find a heuristic solution. • Using real world workload traces from parallel computing cluster for experimentation. [ABSTRACT FROM AUTHOR]

Virtualization technology allows service providers to operate data centers in a cost-effective and scalable manner. The data center network (substrate network) and the applications executed in the data center (virtual networks) are often modeled as graphs. The nodes of the graphs represent (physical or virtual) servers and switches, and the edges represent communication links. Nodes and links are annotated with the provided and required resources (e.g., memory and bandwidth). The NP-hard virtual network embedding (VNE) problem deals with the embedding of a set of virtual networks to the substrate network such that (i) all (resource) constraints of the substrate network are fulfilled, and (ii) an objective is optimized (e.g., minimizing the communication costs). The existing two-step highly customizable model-driven virtual network embedding (MdVNE) approach combines model transformation (MT) and integer linear programming (ILP) techniques to solve the VNE problem based on a declarative specification. MdVNE generates element mapping candidates from an MT specification and identifies an optimal and correct embeddings using an ILP solver. In the past, developers created the MT and ILP specifications manually and needed to ensure carefully that both are compatible and respect the problem description. In this article, we present a novel construction methodology for synthesizing the MT and ILP specification from a given declarative model-based VNE problem description. This problem description consists of a metamodel for substrate and virtual networks, additional OCL constraints, and an objective function that assigns costs to a given model. This methodology ensures that the derived embeddings are correct w.r.t. the metamodel and the OCL constraints, and optimal w.r.t. the optimization goal. The novel model-based VNE specification is applicable to different network domains, environments, and constraints. Thus, the construction methodology allows to automate most of the steps to realize a correct and optimal VNE algorithm compared to a hand-crafted VNE implementation. Furthermore, the simulative evaluation confirms that using MT techniques reduces the time for solving the VNE problem considerably in comparison with a purely ILP-based approach. [ABSTRACT FROM AUTHOR]

The Resource Provisioning (RP) and Task Scheduling (TS) issues has become an attractive paradigms in cloud industry, this is due to the increasing demand for the services provided by virtual machines that are structured by physical servers owned by the data centers of cloud service providers (CSPs). In this paper, we propose a new model based on multi-agent system for the RP and TS reducing the cost of energy using Deep Reinforcement Learning DRL. A Quantile Regression Deep Q Network (QR-DQN) algorithm generates an appropriate policy and the optimal long-term decisions. A set of experiments show the efficiency of our proposed scheduling approach and the performance of our task allocation method.. [ABSTRACT FROM AUTHOR]

Copyright of Telecommunication Engineering is the property of Telecommunication Engineering and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Water clarity provides a sensitive tool to examine spatial pattern and historical trend in lakes trophic status. Yet, this metric has insufficiently been explored despite the availability of remotely-sensed data. We used three Secchi disk depth (SDD) datasets for model calibration and validation from different field campaigns mainly conducted during 2004-2018. The red/blue band ratio algorithm was applied to map SDD for lakes (> 1 ha) based on the first SDD dataset, where R2 = 0.79, RMSE = 100.3 cm, rRMSE = 61.9 %, MAE = 57.7 cm. The other two datasets were used to validate the SDD estimation model, which were indicated the model had a stable performance of temporal transferability. The annual mean SDD of lakes were retrieved across China using Landsat top of air reflectance products in GEE from 1984 to 2018. The spatiotemporal dynamics of SDD were analysed at the five lake regions and individual lake scales, and the average, changing trend, lake number and area, and spatial distribution of lake SDDs across China were presented. In 2018, we found that the lakes with SDDs < 2 m accounted for the largest proportion (80.93 %) of the total lakes, but the total area of lakes with SDD between 0-0.5 m and > 4 m were the largest, accounting for 48.28 % of the total lakes. During 1984-2018, lakes in the Tibetan-Qinghai Plateau lake region (TQR) had the clearest water with an average value of 3.32 ± 0.38 m, while that in the Northeastern lake region (NLR) exhibited the lowest SDD (mean: 0.60 ± 0.09 m). Among the 10,814 lakes with SDD results more than 10 years, 55.42 % and 3.49 % of lakes experienced significant increasing and decreasing trends, respectively. At the five lake regions, except for the Inner Mongolia-Xinjiang lake region (MXR), more than half of the total lakes in every other lake region exhibited significant increasing trends. In the Eastern lake region (ELR), NLR and Yungui Plateau lake region (YGR), almost more than 50 % of the lakes that displayed an increase or decrease in SDD were mainly distributed in an area of 0.01-1 km2, whereas that in the TQR and MXR were primarily concentrated in large lakes (> 10 km2). Spatially, lakes located in the plateau regions generally exhibited higher SDD than those situated in the flat plain regions. The dataset can now be accessed through the website of the National Tibetan Plateau Data Center (http://data.tpdc.ac.cn): DOI: 10.11888/Hydro.tpdc.271571. [ABSTRACT FROM AUTHOR]

With the development of smart grid information physical systems, some of the data processing functions gradually approach the edge layer of end-users. To better realize the energy theft detection function at the edge, we proposed an energy theft detection method based on the power consumption information acquisition system of power enterprises. The method involves the following steps. In the centralized data center, K-means is used to decompose a large amount of data into small data and then input and train neural network parameters to realize feature extraction. We design a neural network named DWMCNN, which can extract features from the day, week, and month and can extract more accurate features. In the edge data center, the random forest (RF) algorithm is used to classify the extracted features. The experimental results show that the clustering method accords with the idea of edge computing-distributed processing and improves the operation speed and that the feature extractor has good convergence performance. In addition, compared with the methods based on various classifiers, this method has higher accuracy and lower computational complexity, which is suitable for the deployment of edge data centers. [ABSTRACT FROM AUTHOR]

The cloud services can be received at anytime from anywhere based on the need of the customers. According to the necessity of the customers, the virtualization of technologies is applied to deliver the cloud services accurately. A large amount of data transfers from user to host and hosts to the user, in cloud environment. To pin the virtual machine on an appropriate host and transferring the data is a challenging task. This paper explains DAVmS: Distance Aware Virtual Machine Scheduling Algorithm, which is applied to arrange virtual machines according to its capability, and pin the VMs to the nearest physical host for accessing the cloud services from the adjacent data center of the customer. This introduced virtual machine scheduling algorithm is used to reduce propagation time and enhance the execution process to reduce the response time. The main purpose of the introduced DAVmS: Distance Aware Virtual Machine Scheduling Algorithm is to deliver an enhanced virtual machine provision policy to physical hosts and access the services from the adjacent data center. [ABSTRACT FROM AUTHOR]

The new trend to design a data center network is to deploy switches and two-port servers on the nodes and edges in an interconnection network, the logic graph of which can be abstracted as a compound graph. Node-independent spanning tree (NIST) is an important structure that can be used to design reliable communication protocols, fault-tolerant broadcasting, and secure message distribution, IP fast rerouting, etc in data center networks. Given a compound graph G (Kn), this paper firstly showed that if the diameter of G was d, then the diameter of G( Kn) was either 2d or 2d + 1. Supposed that there were n NIST rooted at an arbitrary node in an n-regular n-node-connected interconnection network G, this paper presented a unified method to construct n NIST in G (Kn) by proposing an efficient algorithm with time complexity 0 (N), where N was the number of nodes . Ana-lysis of NIST on compound graph Qn(Kn) shows that the maximal height is only the diameter plus 3 . Also, simulation experiments on the augmented based on data center networks show the correctness from another aspect. [ABSTRACT FROM AUTHOR]

Cloud computing is well-known for delivering hosted services over the Internet. These services are mainly divided into three, such as Infrastructure-as-a-Service (IaaS), Platform-as-a-Service (PaaS), and Software-as-a-Service (SaaS). Cloud computing uses the computer system resources, especially storage, and computing power, and is available on-demand without direct active management by the user. The term is generally used to describe data centers available to many users over the Internet. Large clouds are predominant today, and data de-duplication is one of the strategies used to solve the repetition of information. This paper proposes preventing the replication of documents and media records like pictures using the TinEye algorithm. The deduplication strategies are commonly used in the cloud server to decrease the server's gap. To save the unauthorized use of statistics and gaining access to and creating reproduction information on the cloud, the encryption approach is to encrypt the facts before being stored on a cloud server. In this proposed system CloudMe is used for cloud storage. All files are encrypted using the AES algorithm and saved in the real cloud. A secure system architecture layout provides secure deduplication. The design permits the cloud with the critical de duplication functionality to completely cast off the bigger storage and bandwidth price. Thus, the proposed system involves image files in opposition to deduplication strategies to eliminate more storage and bandwidth value. [ABSTRACT FROM AUTHOR]

Cloud computing is a rapidly growing technology that has been implemented in various fields in recent years, such as business, research, industry, and computing. Cloud computing provides different services over the internet, thus eliminating the need for personalized hardware and other resources. Cloud computing environments face some challenges in terms of resource utilization, energy efficiency, heterogeneous resources, etc. Tasks scheduling and virtual machines (VMs) are used as consolidation techniques in order to tackle these issues. Tasks scheduling has been extensively studied in the literature. The problem has been studied with different parameters and objectives. In this article, we address the problem of energy consumption and efficient resource utilization in virtualized cloud data centers. The proposed algorithm is based on task classification and thresholds for efficient scheduling and better resource utilization. In the first phase, workflow tasks are pre-processed to avoid bottlenecks by placing tasks with more dependencies and long execution times in separate queues. In the next step, tasks are classified based on the intensities of the required resources. Finally, Particle Swarm Optimization (PSO) is used to select the best schedules. Experiments were performed to validate the proposed technique. Comparative results obtained on benchmark datasets are presented. The results show the effectiveness of the proposed algorithm over that of the other algorithms to which it was compared in terms of energy consumption, makespan, and load balancing. [ABSTRACT FROM AUTHOR]

With the rapid development of data centers, the energy consumption brought by more and more data centers cannot be underestimated. How to intelligently manage software-defined data center networks to reduce network energy consumption and improve network performance is becoming an important research subject. In this paper, for the flows with deadline requirements, we study how to design the rate-variable flow scheduling scheme to realize energy-saving and minimize the mean completion time (MCT) of flows based on meeting the deadline requirement. The flow scheduling optimization problem can be modeled as a Markov decision process (MDP). To cope with a large solution space, we design a DDPG-EEFS algorithm to find the optimal scheduling scheme for flows. The simulation result reveals that the DDPG-EEFS algorithm only trains part of the states and gets a good energy-saving effect and network performance. When the traffic intensity is small, the transmission time performance can be improved by sacrificing a little energy efficiency. [ABSTRACT FROM AUTHOR]

Climate change is profoundly affecting nearly all aspects of life on earth, including human societies, economies, and health. Various human activities are responsible for significant greenhouse gas (GHG) emissions, including data centers and other sources of large‐scale computation. Although many important scientific milestones are achieved thanks to the development of high‐performance computing, the resultant environmental impact is underappreciated. In this work, a methodological framework to estimate the carbon footprint of any computational task in a standardized and reliable way is presented and metrics to contextualize GHG emissions are defined. A freely available online tool, Green Algorithms (www.green‐algorithms.org) is developed, which enables a user to estimate and report the carbon footprint of their computation. The tool easily integrates with computational processes as it requires minimal information and does not interfere with existing code, while also accounting for a broad range of hardware configurations. Finally, the GHG emissions of algorithms used for particle physics simulations, weather forecasts, and natural language processing are quantified. Taken together, this study develops a simple generalizable framework and freely available tool to quantify the carbon footprint of nearly any computation. Combined with recommendations to minimize unnecessary CO2 emissions, the authors hope to raise awareness and facilitate greener computation. [ABSTRACT FROM AUTHOR]

Efficiency in cloud servers' power consumption is of paramount importance. Power efficiency makes the reduction in greenhouse gases establishing the concept of green computing. One of the beneficial ways is to apply power-aware methods to decide where to allocate virtual machines (VMs) in data center physical resources. Virtualization is utilized as a promising technology for power-aware VM allocation methods. Since the VM allocation is an NP-complete problem, we use of evolutionary algorithms to solve it. This paper presents an effective micro-genetic algorithm in order to choose suitable destinations between physical hosts for VMs. Our evaluations in simulation environment show that micro-genetic approach provides invaluable improvements in terms of power consumption compared with other methods. [ABSTRACT FROM AUTHOR]

Modern data center networks for a fat-tree topology typically adopt a multirooted hierarchical tree structure to achieve multiple-path capability and increase bisection bandwidth. However, the performance of a data center network highly depends on the routing protocols. Conventional routing protocols are unsuitable for modern data center topologies because they lack multiple-path routing support. Another crucial concern in data center networks is load balancing. Certain routing protocol limitations could lead to overloaded or underloaded utilization of a link, thereby considerably reducing the performance of a data center network. Therefore, we present a genetic algorithm (GA)-based dynamic load-balancing routing algorithm, which is heuristic and involves the use of a centralized scheduling technique. This algorithm mainly uses a GA to search for optimal solutions. We implement our algorithm in an OpenFlow controller RYU and Mininet emulator, which is based on software-defined networking architecture. Our evaluation results revealed that our algorithm can effectively achieve load balancing and increase bisection bandwidth. [ABSTRACT FROM AUTHOR]

Workload consolidation is an important method for the efficient operation of cloud data centers, impacting important quality attributes such as resource utilization and power consumption. Many different approaches have been proposed for workload consolidation, but few comparative studies were executed to date. Therefore, it is unclear which of the proposed approaches work best in which situation. In this article, we present a comprehensive simulation‐based comparison of five workload consolidation techniques. We introduce a general framework for workload consolidation techniques to the DISSECT‐CF simulator to foster the development and comparison of efficient data center consolidation algorithms. We use this framework to evaluate the effectiveness of a first fit best fit decreasing heuristic, a custom heuristic, and three population‐based metaheuristics (genetic algorithm, artificial bee colony, and particle swarm optimization). The evaluation is based on a wide variety of real‐world workload traces. The five algorithms are compared in terms of total energy consumption, the duration of the simulation, and the number of migrations. Based on the results, there is no generally best consolidation technique. The results deliver insight into the pros and cons of the algorithms as well as the impact of different parameters. In particular, the results show that population‐based metaheuristics do not offer a significant gain in terms of solution quality to compensate for the increased simulation time. [ABSTRACT FROM AUTHOR]

Cloud computing provides infinite resources and a suitable environment for the execution of large scale computing applications. However, it is also susceptible to frequent failures which can affect users as well as service providers adversely. Therefore, fault tolerance techniques are necessary for the reliable execution of applications in the cloud. This work presents checkpointing based fault tolerance protocols for two types of distributed applications. The first kind of applications is the Bags of Tasks (BoT) applications where an application comprises of a set of independent tasks that do not communicate with each other during execution. Hence, an uncoordinated checkpointing algorithm is proposed for fault tolerance of BoT applications. Subsequently, we consider large scale distributed applications composed of multiple tasks dependent on each other due to inter-task message passing. An uncoordinated checkpointing and message logging protocol is presented for this type of applications. The proposed protocols utilize storage at edge switches in a data center to reduce the bandwidth consumption for saving checkpoints and message logs. Simulation results have demonstrated that the proposed protocols provide an increased rate of successful recoveries from failures and cause lower resource overhead than other contemporary and related schemes. [ABSTRACT FROM AUTHOR]

Aerosol particles affect the Earth’s radiative balance and represent one of the largest uncertainties in climate research. The removal of clouds is the first step and critical in the aerosol retrievals. However, the cloud detection is still challenging. Here, a novel simplified cloud detection algorithm (SCDA) is proposed to identify the cloud and clear-sky over land and based on a simplified radiative transfer model (RTM). The fewer input bands, dynamic thresholds, and only one parameter to be modified are the main advantages of the algorithm, which can be applied to different satellite sensors. In this article, we apply SCDA to the Himawari-8 data in 2016 for preliminary analysis. The detection results are validated using Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observation (CALIPSO) vertical feature mask (VFM) data and the National Centers for Environmental Information (NCEI) ground-based observation data. We also compare the results with the Himawari-8 cloud products from the Japan Aerospace Exploration Agency (JAXA). Compared with CALIPSO VFM data and NCEI ground-based observation data, the correct rate of SCDA cloud detection result is 86.08% and 79.86%, which are higher than that of Himawari-8 cloud products (85.71% and 78.89%). The correct rate of SCDA clear-sky detection result is 88.33% and 87.85%, which are close to the correct rate of Himawari-8 clear-sky products (90.54% and 88.63%). The overall performance of the SCDA is comparable to that of the threshold method for JAXA Himawari-8 cloud products. Therefore, the SCDA can provide accurate cloud mask with only one threshold to be modified and few input parameters. [ABSTRACT FROM AUTHOR]

In recent years, there has been a large increase in the number of cloud users as it provides an easy and exible way to manage user data and applications. With the emerging technologies such as the Internet of things, cloud being the backbone, the load on the cloud servers has increased. The cloud data centers consist of servers hosting multiple virtual machines. One of the main challenges in cloud computing is to efficiently distribute the user service requests to different virtual machines in order to reduce the request processing time and to provide more user satisfaction. Load balancing algorithms basically address two issues: the selection of the data center and the distribution of the load on different virtual machines. Data center selection is handled by service broker policy and distribution of the load on virtual machines handled by VM load balancer. Load balancing techniques play a vital role in minimizing the response time and maximizing throughput and also ensures scalability and reliability. Hence, it has become an important research topic in the field of cloud computing. Cloud-analyst, a java based open source toolkit, is useful to simulate and analyse the load balancing algorithms. In this paper, a comparative study on different service broker policies and VM load balancing algorithms for cloud computing is presented with simulation results. The aim of this comparative study is to find the performance of different service broker policies and the load balancing algorithm tested under different scenarios. [ABSTRACT FROM AUTHOR]

The n -dimensional HSDC with the logic graph H n is one of the most attractive server-centric data center networks for high incremental scalability. The routing design in network topology is very important for information transmission. In particular, the application of disjoint path covers can solve various problems such as program code optimization and mapping parallel programs to parallel structures. H n consists of 2 n cliques and each two vertices in a clique are adjacent. For any vertex sets { x , u } and { y , v } in W and B , respectively, and any two disjoint vertex sets A 1 , A 2 in some clique, where (W , B) is a bipartition of H n , in this paper, we prove that, when | A i | ≠ 1 and A i ∩ { x , y , u , v } = ∅ for i = 1 , 2 , if | A 1 ∪ A 2 | ≤ n − 3 , then there exists an (x , y) -path P 1 and a (u , v) -path P 2 in H n passing through A 1 and A 2 , respectively, such that P 1 and P 2 have no common vertices and P 1 ∪ P 2 contains all vertices of H n. Then P 1 ∪ P 2 is called the 2-disjoint path cover in H n and we propose the efficient algorithms for the path design. Furthermore, we prove that our main result is optimal. • In the problem of data harvest and information transmission, it is very important to design the disjoint path cover in the logic graph of Data Center Network. • We consider the construction of 2-disjoint path cover such that each path should contain all prescribe vertices which have low probability of failure. • The efficient algorithms have been given and we prove that the number of prescribed vertices is not more than n − 3 and explain that this upper bound is sharp. [ABSTRACT FROM AUTHOR]

In hybrid datacenter networks, wired connections are augmented with wireless links to facilitate data transfers between racks. The usage of mmWave/FSO wireless links enables dynamic bandwidth/capacity allocation with extremely small reconfiguration delay. Also, on-demand workload distribution, where the workload of a job is divided into multiple tasks that can be distributed/routed to different racks to be processed in parallel, allows better utilization of computational resources in data centers. In prior work, the dynamic wireless capacity augmentation and workload distribution decisions were mostly made independently and in a heuristic manner for serving distributed and parallel computing jobs. In this paper, we propose a novel analytical framework and algorithms to jointly optimize both the wireless capacity augmentation and the workload distribution, to minimize the job completion time. We consider workload that is not amenable to pipelining, fully amenable to pipelining, and partially amenable to pipelining. With extensive simulation studies, we show that the gain (in terms of the reduction in the job completion time) can be very substantial when allowing such joint optimization. [ABSTRACT FROM AUTHOR]