RESOURCE ALLOCATION MECHANISM BASED ON THE RANKING ON CLOUD ENVIRONMENT FOR SERVICE ACCESS

V.Malarvizhi1 S.Gokilavani2
Angel College of Engineering And Technology Tiruppur
E-mail:42.malar@gmail.com E-mail:mailme.gokilavani@gmail.com ABSTRACT In cloud environment, the optimal allocation of the resources is considered in this paper. The customer can access the services based on their resources needs. But the resource allocation and ranking is difficult in heterogeneous environment . In this paper we proposed improved priority based allocation algorithm for making resource allocation easier with dynamic heuristic mechanism and giving ranking based on the priority .The priority based allocation algorithm mainly for making optimal allocation of resource and mutual relationship between task for accessing the services optimally and dependent task resource allocation problem is solved with dynamic heuristic mechanism. Keywords: resource allocation , dynamic heuristic, service optimality, dependent task resource allocation

Introduction Cloud computing as a new network application mode is springing up. It constructs computing service centre and hires the computing power and storage by using virtualization technology. it divides computing power and storage quickly and freely from cloud to user on demand through network. It transforms all kinds of manufacturing, simulation, and computing resources and abilities into manufacturing services to form a hug manufacturing cloud and distributes them to user the on demand .The challenge is resource allocation and ranking and appropriate service selection for accessing is difficult .we have to meet Quality of service (Qos) parameters also. Resource Allocation Strategy (RAS) is all about integrating cloud provider activities for utilizing and allocating scarce resources within the limit of cloud environment .It requires amount of resources to each application in order to complete a user task optimally. In existing heterogeneous environment various heterogeneous computing resources are integrated and divided as virtual machines by virtualization and assigned to user on demand for computing and simulation.But the heterogeneous resource allocation and ranking is challenging task various kinds of algorithms and framework available for allocation and ranking. But the optimal service selection and resource allocation based on the ranking is difficult at a time. Cloud resources consist of physical and virtual resources.The physical resources are shared across multiple compute requests through virtualization and provisioning. For that we are implementing new allocation algorithm and dynamic heuristic mechanism. In proposed system we are focusing optimal resource allocation with ranking and proposed priority based allocation algorithm with dynamic heuristic mechanism. A new improved intelligent algorithm is presented to solve OACR based n the ranking a with dynamic heuristic mechanism. Simulation results on standard tests show that this new hybrid algorithm is pretty efficient to solve this kind of high-dimensional complex combinatorial problems. In traditional Service-Oriented Architecture (SOA) is extended from software application to generalized cloud service with the inclusion of both software applications and hardware equipments with good interoperability, selforganization and scalability

Fig: 1 manufacturing process in cloud

 Related Work In this section, we discuss the related work in the areas of heterogeneous resource allocation, QoS and ranking for service access in the cloud environment.. In [1] allocation based on the user needs, In [2,3[ for minimizing the changes in resource allocations due to adjustment ,In [8] authors have explained the algorithm for negotiation protocol for resource provisioning in detail. In [9] authors propose a model and a utility function for location-aware dynamic resource allocation. In [10] A comprehensive comparison of resource allocation policies is covered. In [11] author has used a Genetic Algorithm for scheduling of tasks in cloud computing systems.In [14]According to the characteristics of cloud computing, semantic properties of cloud service are studied, In [15] The classification, management, provision, storage and evaluation of cloud services are investigated widely. Pre-decision and online decision of SCOS are also deliberated in different ways. Proposed System The of proposed system is to optim al allocation of computing resources(OACR) and integrates various cooperating components to achieve high performance optimal resource allocation and ranking. Resource allocation deals with the resource availability information like service access capacity, distribution and allocation of resources. Ranking involves the modelling and analyzing the user priority. Resource allocation service is based on the allocatin capacity and dynamic heuristic mechanism. In cloud computing, two crucial optimization factors in determining resource sharing efficiency and platform application performance are SCOS and OACR exactly. In this paper we proposed priority based allocation algorithm for making allocation easier based on the ranking. Quality of service attributes should be satisfied during the service distribution. The ranking of computing resources totally depends on the corresponding upper layer selected cloud services. Using the distributed characteristics of services and infrastructures, the available computing resources are reduced and the OACR are constrained by the upper layer decision. With the purpose of efficient decision the following three issues are need to be studied (1) QoS parameters of software / hardware cloud services and computing resource respectively and the mutual relation between them; (2)The problem formulation of optimal allocation of resources with dynamic heuristic mechanism for multi-objectives and multi-constraints in

(3)The proficient resource allocation algorithm for addressing large-scale resource allocation problem with dynamic heuristic mechanism
A The Structure Of OACR

The OACR of CMfg is composed of three levels: manufacturing task level, virtual resource level, and computing resource level, as shown in Fig. 4. On the manufacturing ask level, assume the tasks of a given MTs set are meta tasks. Meta task means that the task is inseparable for executing in VMs/CRs, For instance in a multidisciplinary collaborative simulation, each module runs on one VM with users control and interaction. Each VM is inseparably running on only one CR. So MTs and VMs have the one-to-one mapping relationship. When MTs demands are abstracted as virtual resources demands, the virtual machine manager receives the demand information and allocates available VMs for physical manufacturing resources. The physical manufacturing resources can be not only manufacturing/simulation equipments and computing resources also. Manufacturing (or) simulation equipment needs a CR to control and monitor. Thus, all VMs are supported by CRs. They form the virtual resource level and support the running of MTs. Because the customized MTs are applied by user, the constraints of VMs (e.g., the demand of memory size, computing speed, communication link and bandwidth, etc.) could be obtained at the same time., the mapping of manufacturing task level and virtual resource level is the foundation of OACR, and the mapping of virtual resource level and physical resource level is central to the optimization. In this paper, the manufacturing task level and virtual resource level are merged, and the optimal allocation of MTs (or VMs) and CRs under the concrete computation and communication constraints is emphasized.

Fig: 2 resource allocation framework

B Review Of GA, ACO, AND IA GA is an adaptive global optimization stochastic search algorithm which is inspired by the principle of evolution and natural genetics. With a set of structured populations which represented the candidate solutions of the problems, it combines the survival of the fittest among populations (selection), the structured yet randomized information exchange (crossover), and the random bit mutation Firstly, roulette wheel selection is used commonly in the standard GA. It is performed by randomly picking a certain amount of populations according to their fitness values to form a new group of populations. The one with higher fitness value occupied higher probability of selected. Secondly, each two of the selected populations exchange parts of their gene bits in the crossover operation. So two new genetic chromosomes are generated, and the better gene bits go into the next generation. Thirdly, the mutation operation randomly changes some gene bits of populations with a certain probability for increasing the diversity.ACO is a kind of swarm intelligence algorithm which takes inspiration from the social behaviours of ant colony. It combines ants routing and pheromone update. The original intention of ACO is to solve the complicated path optimization problems, such as TSP, and edge scheduling. In the process of routing and finding foods, ants deposit pheromone on the path they have walked in order to mark some favourable path and broadcast the information. The longer the path, the lower the density of pheromone is. Other ants can perceive this pheromone and recognize its density. They have a large probability to select the path which has the greater pheromone density. Then a kind of information positive feedback is formed. The pheromone density on the optimal path will become higher, while the pheromone density on other paths will reduce as the time goes by. Finally the whole colony will find the optimal path. IA is a kind of evolutionary programming which based on the immune system in biotic science. With the introduction of the concepts and the characteristics of antige recognition,immunological memory, and immune regulation, diversified immune algorithms are presented. IA is a convergence of immune theory and genetic algorithm. It contains genetic progress, immune inoculation, and immune selection, but first of all, antigen extract and vaccine selection according to the feature information of problem is the most important part of this algorithm. It is a core rule to lead the population evolution in the right direction. Then, the population initialization and the genetic evolution are all the same as the standard genetic algorithm. After selection, crossover, mutation, new populations are vaccinated by antibodies. It is injecting a priori knowledge into the new populations in some degree for improving their fitness values (the a priori knowledge in IA is the same as in ACO.) Then, new populations are selected by immune selection operation according to the choosing rules of simulated annealing. C Proposed Priority Based Ranking Algorithm In this proposed algorithm mainly for allocating resources on heterogeneous environment. In previous many algorithm is used for resource allocation. But its only for resource allocation locally. So improved priority based ranking algorithm is proposed.

In this proposed algorithm mainly for allocating resources on heterogeneous environment. In previous genetic algorithm is used for resource allocation. But its only for resource allocation locally. So priority based resource allocation algorithm is proposed. The algorithm determines the initial location and the coding scheme of population . In this we using allocation algorithm for selecting people and allocating resources. The allocation operator having two bits for service. service-genebit which represents the selected cloud services and resource-gene bit which represents the allocated computing resources are needed to be set. One task corresponds to two gene bits. When a set of tasks are submitted to cloud system, the system should choose suitable cloud services and computing resources with specific serial numbers at the same time. Assume the length of gene code be twice of the number of tasks. Each two genebits represent the serial number of the selected service and the allocated computing resource for the corresponding task. It briefly demonstrates the relationship between cloud service and computing resource and makes the optimal process more convenient.

The collection of tasks in cloud computing location can be presented as a directed acyclic graph (DAG G ={N; W; U; Ht; Hc; He; Ht}. Where: The set N = {Ni}i = 1 : n represents tasks with serial numbers, where n is the total number of tasks. The set W = {Wi}i = 1 : n indicates the size of tasks. The set U = {Uji}i =1 : n; j =1 : n represents the communication relationships among tasks, where Uji reflects the communication from task Ni to task Nj. We should note that UijUji . If there is no communication between the two tasks, then Uji = 0. Ht ={Hi(i)i 1 : n,},{ Hc =Hc(i)i = 1 : n} He ={He(i)i = 1 : n} and Hr ={Hr(i)i =1 : n represent the lowest time, cost, energy and reliability requirements of tasks respectively. Besides, let the predecessor tasks set of Ni be Li, and the successor tasks set be Ri. The node with no predecessor task Li = is named source node, and the node with no successor task Ri= is called sink node. All tasks strictly observe the tasks, priority rules, a node can only be started after all output communication data of its predecessor household tasks are obtained. If the amount of submitted task is N, the total execution time T, the total cost C and the total energy-consumption E of the software and hardware service can be calculated as follows respectively.
For software services, N T= S cN C=Tc= s For hardware services N Nc +s + =N S s +s E= Te=eN s (4)

Select N individuals according to 1, 2 ......k and generate new I

E Dynamic heurictic algorithm

Dynamic_Heuristic (I) For (i = 1 N) r = Ii Randomly choose a genebit k[1 : M] If (p corresponds to service-bit) Randomly choose hns values from 1 to ns Choose the service si with the highest heuristic Ys =maxj Ys(,j)( j [1, hns_) Gr(k) = si Else Randomly choose hnp values from 1 to ns Decide the computing resource pi with the maximum heuristic Yp = maxj yp( j)( j 1,hnp_) Gr(k)=pi End if Agree to r with simulation annealing possibility End for

(1)

(2)

Compared with other algorithms, this priority based allocation algorithm for the dynamic vaccination and pheromone updating strategy are added in this. Allocation algorithm is used for improving exploration during searching, dynamic vaccination and pheromone updating strategy is taken for further improving the exploitation and searching direction with the dynamical consideration of both computation and communication in resource allocation with ranking. The genetic evolution just adopts the standard roulette wheel strategy, singlepoint crossover and mutation.

T=

(3)

DAG 1

D Resource Allocation Algorithm

Resource_allocation (I) identify the selection range according to Pselection Sort one with QUICK sort algorithm and stored as V Estimate the number of times of I1 to be selected, 1 =1+ (2N/K _ 2)Fa=F1 Calculate = 2(N _ 2K)Fa/K(K _ 1)F1 Calculate 1, 2, 3..... k other K _ 1 individuals

DAG 2

VI CONCLUSION For solving the new complex model, an priority based allocation algorithm is presented with the introduction of other three most commonly used intelligent algorithms (i.e., GA, ACO, IA). By the introduction of efficient strategies, Priority based allocation algorithm showed high performances in terms of searching ability, stability, and time complexity in solving the OACR problem compared with the other three algorithms. At the same time, extensive experiments demonstrated the design of heuristic information in OACR and the suitable heuristics in different situations. VII REFERENCES [1] Yuanjun Laili &,Fei Tao , Lin Zhang &,Bhaba R. Sarker A study of optimal allocation of computing
resources in cloud manufacturing systems Published on Elsevier journal 2012

DAG 3

Fig: 4 DAG FOR ROUTE SELECTION

IV EXPERIMENTS AND DISCUSSION Thus a composite project can be divided into far more than 5 tasks. Those tasks have several functional and non-functional requirements for cloud services. A few of them need hardware cloud services, some need software cloud services. In order to evaluate the performance of priority based optimization compared with the others decision .we use basic real-coding GA uniformly to simulate the decision process in theory. At the OACR step, each gene-bit represents the selected computing resource number for the selected service. The lengths of gene-bits having two steps. For testing the OACR models three DAG graph is introduced. In the experiments, all of the communication costs and the computation costs are randomly generated. Due to the looser communication time constraints, the experiments focus mainly on the effect of the computing speed, memory, and reliability constraints of CRs. The priority based allocation algorithm proposed with QOS and other algorithm was used for locally and globally in resource allocation .From the global perspective, priority based allocation algorithm showed high performance in all scales of MTs in OACR. Priority based allocation algorithm exploration, and the introduction of experiential pheromone and dynamic heuristics improved the algorithms exploitation.priority based allocation algorithm got a better balance between exploration and exploitation by these two strategies for addressing OACR in CMfg. From the perspective of solution quality and stability, priority has big potential in solving this kind of allocation problem without the increase of time complexity V FUTURE ENHANCEMENT Optimal resource allocation and ranking is improved efficiently with quality of service attributes and dynamic heuristic mechanism. In future reservation mechanism and framework was introduced for avoid the collusion during the allocation using intelligent algorithm

[2] M. Aron, P. Druschel, W. Zwaenepoel, Cluster reserves a mechanism for resource management in cluster-based network servers, in: Proc. of the ACM Sigmetrics International Conference on Measurement and Modeling of Computer Systems, June 2000. [3] Bo An, Victor Lesser, David Irwin and Michael Zink, Automated Negotiation with Decommitment for Dynamic Resource Allocation in Cloud Computing, Conference at University of Massachusetts, Amherst, USA.
[4] F. Tao, D. Zhao, Y.F. Hu, Z.D. Zhou, Correlation-aware resource service composition and optimal-selection in manufacturing grid, European Journal of Operational Research 201 (1) (2010) 129143. [5] K. Fujii, T. Suda, Semantics-based dynamic service composition, IEEE Journal on Selected Areas in Communications 23 (12) (2005) 23612372.

[8] N. Bansal, A. Caprara, M. Sviridenko, Improved approximation algorithms for multidimensional bin packing problems, in: Foundations of Computer Science, 2006, pp. 697_708. [9] Gihun Jung and Kwang Mong Sim, Location-Aware Dynamic Resource Allocation Model for Cloud Computing Environment, International Conference on Information and Computer Applications(ICICA), IACSIT Press, Singapore, 2012. [10] Chandrashekhar S. Pawar and R.B. Wagh, A review of
resource allocation policies in cloud computing , World Journal of Science and Technology, 2(3):165-167, 2012. [11] Sandeep Tayal, Tasks SchedulingOptimization for the Cloud Computing systems, International Journal of Advanced EngineeringSciences and Technologies (IJAEST), 5(2): 111 115, 2011. [12] Daniel Warneke and Odej Kao, Exploiting dynamic resource allocation for efficient parallel data processing in the cloud, IEEE Transactions On Parallel And Distributed Systems, 2011.

[13] Atsuo Inomata, Taiki Morikawa, Minoru Ikebe and Md. Mizanur Rahman, Proposal and Evaluation of Dynamic Resource Allocation Method Based on the Load Of VMs on IaaS, IEEE, 2010. [14] K. Fujii, T. Suda, Semantics-based dynamic service composition, IEEE Journal on Selected Areas in Communications 23 (12) (2005) 23612372. [15] A.J. Ferrer, F. Hernandez, J. Tordsson, E. Elmroth, A. AliEldin, C. Zsigri, R. Sirvent, J. Guitart, R.M. Djemame, W. Ziegler, T. Dimitrakos, S.K. Nair, G. Kousiouris, K. Konstanteli, T. Varvarigou, B. Hudzia, A. Kipp, S. Wesner, M. Corrales, N. Forgo, T. Sharif, C. Sheridan, OPTIMIS: a holistic approach to cloud service provisioning, Future Generation Computer Systems 28 (1) (2012) 6677.