An Energy Aware Resource Utilization Framework to Control Traffic in Cloud Network and Overloads

Energy consumption in cloud computing occur due to the unreasonable way in which tasks are scheduled. So energy aware task scheduling is a major concern in cloud computing as energy consumption results into significant waste of energy, reduce the profit margin and also high carbon emissions which is not environmentally sustainable. Hence, energy efficient task scheduling solutions are required to attain variable resource management, live migration, minimal virtual machine design, overall system efficiency, reduction in operating costs, increasing system reliability


INTRODUCTION
Cloud is collection of interconnected virtualized dynamically provisioned on demand computing resources based on pay-as-you-go model [1]. The energy consumption and resource utilization are coupled as high energy consumption in cloud is due to the low utilization of computing resources as compare to efficient utilization of computing resources. As per the studies, the average resource utilization in most of the data centers is lower than 30% [2], and the energy consumption of idle resources is more than 70% of peak energy [3]. This massive energy consumption causes significant CO2 emissions, as many data centers are backed by "brown" powerplants.
Cloud data centers are electricity guzzlers especially if resources are permanently switched on even if they are not used. An idle server consumes about 70% of its peak power [4]. This waste of idle power is considered as a major cause of energy inefficiency.
This paper makes a study on taske scheduling policies for energy efficiency and propose an energy aware task scheduling algorithm based on cache memory and broadcasting.
Rest of the paper is organized as follows. Section 2 investigates previous research in energy aware techniques. Section 3 presents a novel energy aware resource utilization framework to control traffic in cloud networks and overloads. Finally, Section 4 concludes the paper. Table 1 represents the study of previous energy aware techniques. Step Towards Green Computing [6] Proposed an architectural principle for energy efficient management of Clouds, energy efficient resource allocation strategies and scheduling algorithm considering Quality of Service (QoS) outlooks.

QoS
CloudSim Results show that this approach is effective in minimizing the cost and energy consumption of cloud applications thus moving towards the achievement of Green Clouds. Energy Efficient Scheduling of HPC Applications in Cloud Computing Environments [7] Proposed a near-optimal scheduling policy that exploits heterogeneity across multiple data centers for a Cloud provider. Also examined how a Cloud provider can achieve optimal energy sustainability of running HPC workloads across its entire Cloud infrastructure.  [12] In this paper, the problem of global operation optimization in cloud computing is considered from the perspective of the cloud service provider (CSP) to provide a versatile scheduling and optimization framework that aims to simultaneously maximize energy efficiency and meet all user deadlines, which is also powerful enough to handle multi-user large scale workloads in large scale cloud platforms. Proposed a cooperative two-tier energy aware scheduling technique to establish a constructive cooperation between the schedulers of a broker and its hosts in order to reach an optimal scheduling in terms of power consumptions and turnaround time.
Power consumption, turnaround time.

CloudSim
Results show that the proposed task scheduling approach not only reduces the total energy consumption of a cloud by 41%, but also has profound impacts on turnaround times of real-time tasks by 85%. Proposed a DVFS-enabled energyefficient workflow task scheduling algorithm DEWTS in order to obtain more energy reduction as well as maintain the quality of service by meeting the deadlines.
Energy consumption ratio (ECR), system resource utilization ratio, average execution time, energy saving ratio.

CloudSim
Results show that DEWTS can reduce the total power consumption by up to 46.5 % for various parallel applications as well as balance the scheduling performance.
An Energy-Saving Task Scheduling Strategy Based on Vacation Queuing Theory in Cloud Computing [21] In this the average sojourn time of tasks and the average power of compute nodes in the heterogeneous cloud computing system under steady state is analyzed. Next, based on the busy period and busy cycle under steady state, the expectations of task sojourn time and energy consumption of compute nodes in the heterogeneous cloud computing system is analyzed and energy consumption is reduced.

Matlab
Results show that the proposed algorithm can reduce the energy consumption of the cloud computing system effectively while meeting the task performance.
A novel energy-efficient resource allocation algorithm based on.
The objective of this paper is to optimize resource allocation using an improved clonal selection algorithm (ICSA) based on makespan optimization and energy  [22] consumption models in cloud computing environment.
A new energy-aware task scheduling method for data-intensive applications in the cloud [23] In this method, first, the datasets and tasks are modeled as a binary tree by a data correlation clustering algorithm, in which both the data correlations generated from the initial datasets and that from the intermediate datasets have been considered. Hence, the amount of global data transmission can be reduced greatly, which are beneficial to the reduction of SLA violation rate. Second, a "Tree-to-Tree" task scheduling approach based on the calculation of Task Requirement Degree (TRD) is proposed, which can improve energy efficiency of the whole cloud system by reducing the number of active machines, decreasing the global time consumption on data transmission, and optimizing the utilization of its computing resources and network bandwidth Network bandwidth, resource utilization Results show that the power consumption of the cloud system can be reduced efficiently while maintaining a low-level SLA violation rate.
DENS: data center energy efficient network aware scheduling [24] This work underlines the role of communication fabric in data center energy consumption and presents a scheduling approach that combines energy efficiency and network awareness, named DENS to balance the energy consumption of a data center, individual job performance, and traffic demands. The proposed approach optimizes the tradeoff between job consolidation and distribution of traffic patterns.
Energy efficiency and network awareness individual job performance, and traffic demands CloudSim A novel virtual machine deployment algorithm with energy efficiency in cloud computing [25] To improve the energy efficiency of large-scale data centers, TESA is firstly proposed. Then based on the TESA, five kinds of VM selection policies are presented. Considering energy efficiency, the MIMT is chosen as the representative policy to make comparison with other algorithms.

CloudSim
Results show that, as compared with single threshold (ST) algorithm and minimization of migrations (MM) algorithm, MIMT significantly improves the energy efficiency in data centers.
Energy efficient scheduling of virtual machines in cloud with deadline constraint [26] Proposed an energy efficient scheduling algorithm, EEVS, of VMs in cloud considering the deadline constraint, and EEVS can support DVFS well. A novel conclusion is conducted that there exists optimal frequency for a PM to process certain VM, based on which the notion of optimal performance-power ratio is defined to weight the homogeneous PMs.  [28] In this paper, two energy-conscious task consolidation heuristics are presented.
These heuristics maximizes the resource utilization and explicitly takes into account both active and idle energy consumption. It assigns each task to the resource on which the energy consumption for executing the task is explicitly or implicitly minimized without the performance degradation of that task.

Energy consumption, resource utilization
The results in this study should not have only a direct impact on the reduction of electricity bills of cloud infrastructure providers, but also imply possible savings (with better resource provisioning) in other operational costs (e.g., rent for floorspace).

EnergyAware
Genetic Algorithms for Task Scheduling in Cloud Computing [29] In this paper independent tasks scheduling in cloud computing as a biobjective minimization problem is considered with makespan and energy consumption as the scheduling criteria. Dynamic Voltage Scaling (DVS) is used to minimize energy consumption and to propose two algorithms to find the right compromise between make span and energy consumption.

Makespan, energy consumption
Results show that the two algorithms can efficiently find the right compromise between make span and energy consumption.

Energy Efficient Migration and Consolidation
Algorithm of Virtual Machines in Data Centers for Cloud Computing [30] In this a dynamic energy efficient virtual machine migration and consolidation algorithm based on a multi resource energy efficient model is proposed. This algorithm has minimized energy consumption with Quality of Service guarantee and also reduced the number of active physical nodes and the amount of VMs migrations.
Energy consumption, quality of service Results show better energy efficiency in data center for cloud computing Energy-Efficient Scheduling of Urgent Bag-of-Tasks Applications in Clouds through DVFS [31] In this a cloud aware scheduling algorithm is proposed that applies DVFS to enable deadlines for execution of urgent CPU-intensive Bag-of-Tasks jobs to be met with reduced energy expenditure. Algorithm has significantly reduced the energy consumption of the cloud while not incurring any impact on the Quality of Service offered to users.

CloudSim
Results show that proposed approach reduces energy consumption with the extra feature of not requiring virtual machines to have knowledge about its underlying physical infrastructure.

Enhanced
Energyefficient Scheduling for Parallel Applications in Cloud [32].

An
Enhanced Energy-aware Scheduling (EES) heuristic algorithm is proposed to reduce energy consumption still meeting performance based SLA in data center running parallel applications. This algorithm ensures that the job finish before the deadline decided at the earliest. The main idea of this approach is to study the slack room for the non-critical jobs and try to schedule the tasks nearby running on a uniform frequency for global optimality.  [34] In this an energy-aware online provisioning approach is proposed for HPC applications on consolidated and virtualized computing platforms. Energy efficiency with an acceptable QoS penalty is achieved using a workload-aware, just-right dynamic provisioning mechanism and the ability to power down subsystems of a host system that are not required by the VMs mapped to it. Proposed SEATS, a virtual machine scheduling algorithm, which aims to reach the optimal level of utilization by offering more computing power to virtual machines of a host. SEATS makes hosts execute their virtual machines faster to reach their optimal utilization levels without needing to migrate virtual machines which eventually leads to reducing power consumption.

CloudSim
Results show that proposed method not only reduces total energy consumption of a Cloud by 60 %, but also has a profound impact on turnaround times of real-time tasks by 94 %. It also increases the acceptance rate of arrival tasks by 96 %.
Real-Time Tasks Oriented Energy-Aware Scheduling in Virtualized Clouds [38].
To guarantee system schedulability a novel rolling-horizon scheduling architecture is proposed for realtime task scheduling in virtualized clouds. Then a task-oriented energy consumption model is given and analyzed. Based on scheduling architecture, a novel energy-aware scheduling algorithm named EARH is proposed. System schedulability, energy consumption.

CloudSim
Results show that EARH significantly improves the scheduling quality of others and it is suitable for real-time task scheduling in virtualized clouds.
Furthermore, two strategies are proposed in terms of resource scaling up and scaling down to make a good trade-off between task's schedulability and energy conservation. Proactive Scheduling in Cloud Computing [39] Proposed a service ranking algorithm on the basis of detailed performance monitoring and historical analysis and based on their contribution, a weight age is assign to all service quality factors or performance metrics and as a

PROPOSED ENERGY AWARE RESOURCE UTILIZATION FRAMEWORK
Cloud Computing is one of the most fast evolving computing platform which is the future of supercomputing. A time will come when everyone would be on the cloud network and at that time it is essential for the cloud network to perform well. Cloud is also a computing server and hence it takes every order as million instruction set. These instruction sets are often referred as Jobs. Scheduling an instruction set or job requires a lot of computing one wrong placement may lead to wastage of energy units. The proposed work has taken these issues in a very serious manner and has designed an architecture diagram which deal with the job scheduling process from start to end. The proposed algorithm covers placement of the job at server, monitoring of the server to prevent them from overloading and when they are exhausted from jobs, the creation of Virtual Machine is also a part of the proposed work. The proposed algorithm enhances the MBFD Algorithm by introducing artificial intelligence to it.
In this research paper, we are particularly focusing on the cloud server maintenance and scheduling process and to do so, we are using the interactive broadcasting energy efficient computing technique along with the cloud computing server. Job handling has been done using one of the finest swarm intelligence techniques called Artificial Bee Colony Algorithm. Artificial Bee Colony algorithm monitors the performance of the servers or host in order to check that they do not get overloaded. Figure 1 represents a complete transaction process from user to server. The user can have n number of jobs and its request would be posted on central server. The central server looks into the requirement of the user and checks into the cache memory. If any service of such kind is already done in the past and if there are  [40] In this an integrated assessment model considering both resource credibility and user satisfaction is established and a resource scheduling strategy based on genetic algorithm is designed on the basis of this model. Resources credibility, user satisfaction.

CloudSim
The numerical results show that this scheduling strategy improves not only the system operating efficiency, but also the user satisfaction.
Research on Batch Scheduling in Cloud Computing [41] This paper provides the task scheduling algorithm based on service quality which fully considers priority and scheduling deadline. The improved algorithm combines the advantages of Minmin algorithm with higher throughput and linear programming with global optimization, considers not only all the tasks but also the high priority tasks. Balancing Consumption [43] In this an intelligent optimizing strategy of virtual resource scheduling is designed which fully takes into account the advantages of cloud virtual resource. It improves the selection and cross processing in GA and takes the optimal span and load function as the double fitness function to make the resource scheduling efficiency improved obviously. 1025 more than two vendors or sub servers who have done similar kind of work then it goes for the feedback of the subservers. There can be N number of sub server. Here in the architecture diagram it is represented by S1….SN. The sub server takes the tasks from the central server and executes them in timely fashion. If any sub server has a feedback for similar kind of job then the availability of the sub server is checked and if it is available then the work is provided to the sub server. The concept of the broadcast is applied on two places. First when there is no server in the cache memory or in the feedback queue and second when the feedbacks queue server is unavailable. The central server acts back on the responds of the sub servers. The response can be only taken from those sub servers who are in the range of the user demand. The range would be calculated with the help of distance formula. Another situation is considered that the sub server is overloaded with tasks and it is unable to provide memory it to its VMs. In such a case the VMs would have to be migrated from one sub server to another sub server. The process takes a lot of energy if not done efficiently. In order to attain the goal artificial bee colony algorithm is applied. The artificial bee colony algorithm takes the available servers as input bees and process them according to the designed fitness function.

CONCLUSION
The energy efficiency of computing resources plays a significant role in the overall energy consumption of the data center. The energy efficient task scheduling solutions are required to attain variable resource management, live migration, minimal virtual machine design, overall system efficiency, reduction in operating costs, increasing system reliability, and prompting environmental protection with minimal performance overhead. This paper provides a comprehensive overview of the energy efficient techniques and approaches and proposes the energy aware resource utilization framework to control traffic in cloud networks and overloads.