# Introduction he cloud computing is a specialized environment with a huge collection of computer systems which are connected to internet with either public or private networks. This cloud computing environment provides facility to the end user on demand. This technology has reduced the price to acquire the different softwares, operating systems, and tools which were only available to the larger companies upon huge investments. These services are available on demand [1] and the services are metered and the user needs to pay only for the services are used. It is one or other way to customized usage of information technologies. Cloud computing is a collection of heterogeneous services under a single umbrella of cloud service provider. Therefore in many cases we do not need to install or acquire new devices to have some specialized services by the cloud computing service providers such as, Google provides its several type of services to the end users such as Gmail, Google docs, Google maps etc. Microsoft Company is also providing the services such as Microsoft Office 365 and Microsoft azure to the end users. Sales force, Amazon is also providing the services to its consumers with the user needed services. The cloud computing services are provided on demand and as per the user requirements. This process does not need to be done by somebody manually perhaps automatically which is called as software automation process. This type of dynamic provisioning provides not only increased service capacity but also provides reliability, security and privacy. The cloud services can be had using any one of the devices using such as personal computers, mobiles etc using any type of network access. # II. # Computing at data Centre Level The data centre and cloud architectures [4] are keeping on advancing, to address the requirements of expansive scale multi server farms in clouds. These requirements are revolved around seven measurements called (i) scalability, (ii) storage, (iv) bandwidth, (v) speed in network services, (vi) efficiency in memory utilization, (vii) agility in service creation, (viii) cost productivity. # III. Review of Related Literature # T. Kokila vani has proposed Load Balanced Min Min scheduling set of rules [5] which produces higher effects than minmin scheduling set of rules. It reduced the make span and attained improvement. 2. Jens Buysse et al. described a new method to minimize [6] the power intake. This routing and scheduling algorithm selects unused nodes and turns off those nodes. In step one, it chooses the special nodes and which are marked as the destination nodes. In the second step, it decides the unicast route to a particular destination to supply. The results illustrated the power consumption reducing by 50% lesser in comparison with other standard scheduling and routing algorithms. # Wanneng Shu et al. Proposed a new type of power effective algorithm in which the resources [7] are allocated in a cloud computing environment. The CloudSim device kit is used to study the performance of the set of rules. Time, cost and energy intake are used as the parameters in this set of rules. The response time and make span has drastically reduced. 4. M. Hemamalini, M. V. Srinath, et al carried out their stud on a heuristic set of rules for data placement. In this study, the heuristic method focuses [8] on the node scheduling which is used to reduce the energy consumption in the cloud. This study has accommodated the maximum data on the minimum quantity of nodes. It has used the greedy approach for this process. The algorithm is implemented with the use of Cloud Sim. 5. Ms. Nitika, Ms. Shaveta, Mr. Gaurav Raj carried their research on Equally Spread Current Execution Algorithm. In this study, the set of rules are used to handle the system with priorities and it has distributed the burden randomly [9]. This distribution is done with the help of a parameter scale. This process transfers the load to the lightly loaded digital gadget or handles that venture and takes much less time. It is observed by the researchers that it provides the maximize throughput. This set of rules makes use of the spread spectrum method to balance the weight of the task considering into more than one virtual machines. 6. Bhuvnesh Pathania et al. proved the performance of his proposed approach which has reduced the electricity consumption. In this approach virtual machines are loaded with variable load [10]. This load is balanced in the grid clusters which are enough to manipulate the required Quality grade. 7. Jianfeng Zhao and Hongze Qiu et al. provided their research related information with the help of a replication approach to achieve many unique possibilities, such as lowering the system running time and power consumption. The algorithm has [11] employed genetic protocol and used ant colony optimization technique. 8. Javid Taheri et al. has worked on match making scheduling segment and offered algorithms to lessen the make span for executing all jobs and their transfer time [12]. It used two distinct set of rules for scheduling the activity to mirror the information fields to the linked nodes. [15] a new type of and quicker heuristic method. This method tried to solve the hassle of mapping independent scheduling to a virtual machine. IV. # Scheduling Generally, scheduling [16] is a term used for a set of rules that will govern the order by the execution of a particular process. Here the process is a word used for an executable programming part. A process execution will undergo in the computer system such as a CPU (central processing unit) burst followed by an IO (input output) burst again a CPU (central processing unit) burst followed by an IO (input output) burst and so on. 'Burst' is an action with as fast as possible highest speed and continued action. Scheduling is an important property of an operating system. Scheduling process can be looked as 'Service request scheduling' and 'Resource response scheduling'. Generally, service request scheduling occurs due to (i) When the user submits his or her request to the service provider (ii) Service provider executes the request (iii) Processing the request in the service request architecture (iv) Dynamic virtual machine generation and dispatch at the provider site. A computer system needs scheduling before use; CPU (central processing unit) is one of the most critical parts in the computer. Multiprogramming is one of the basic and important scheduling techniques. Generally CPU (central processing unit) scheduling is done in such a way as to keep it busy as much as possible. Theare two issues (i). Jobs must be smaller so that they could receive large fraction of number of processors they requested. (ii). Execution efficiency of smaller jobs may be low. V. # Virtualization In cloud computing, virtualization [17] is a key idea. Virtualization improves the effectiveness of computing resources utilization and additionally the dynamic resource provisioning capacities in a cloud. The virtual machine, whose associated necessities cannot be domestically fulfilled, are selected for migration. This set of rules searches the maximum loaded virtual machines and allocate the load in an effective manner. Virtualization provides the applications to migrate from one server to another. This migration is possible dynamically, which means there is no need to make the server down, and the workload can be easily managed. # VI. # Load Balancing The main objective of load balancing [18] in a computing environment is to guarantee that no single system is overloaded with tasks, while the other physical node is left idle. The key criteria for a good load balancer are to maintain a balanced state of workload among the actively participating computational nodes. Apart from the central task of load balancing, scheduled jobs, it also focuses on maximizing the throughput, minimizing the response time and better resource utilization. A cloud load balancer should have the characteristics in such a way that the cloud service providers will not be overloaded with the set of requests. # VII. Internal Working Mechanism of Hybrid Algorithms The algorithm is a step by step method with a set of rules that can provide solutions within a finite number of steps. A hybrid algorithm [19] is a combination of two or more algorithms. Before a user sends a 'request' to the load balancer, it resolves the load balancer's domain name using a domain name server (DNS). Domain name server returns one or more internet protocol addresses (IP address) to the client machine. With network load balancers, load balancing creates a network interface for each available virtual machines list which is either busy or idle .Each load balancer node in this virtual machine list uses this network interface to get a static internet protocol address (static IP Address). It can be optionally associated one internet protocol address with each network interface when it creates the load balancer. As traffic to the application changes over time, load balancing scales the load balancer and updates the DNS entry. Here one important aspect is that domain name server entry specifies the TTL (Time to live) as sixty seconds. TTL makes the internet protocol address can be re mapped quickly in response to change the traffic. The client machine determines which internet protocol address need to use for sending the requests to the load balancer. Load balancer node requests the server. Load balancer node selects a virtual machine which can handle the request. Now the load balancer node sends the request to the target using its private internet protocol address (Private IP Address). When a load balancer accepts incoming traffic from the client and routes the requests to its virtual machine pool. The load balancer also checks whether the virtual machine is either idle or busy and ensures that it routes the traffic only to the virtual machines which are capable to handle. If it is being used equally spread current execution algorithm, for example, it is cross zone load balancing. In this load balancing, the nodes of the load balancer would distribute the requests regardless of the availability of virtual machines and distributes the traffic evenly across all the virtual machines in the pool of virtual machines. With the application of load balancer, the load balancer node receives the request and evaluates the priority order to determine which rule to apply and then selects a virtual machine from the pool of virtual machines for the rule action using the routing algorithm. Routing is performed independently for each one of the virtual machines in the virtual machine pool, even though when the virtual machine is assigned to the pool of virtual machines. With network load balancers the load balancer node receives the connection to select the virtual machine from the virtual machine pool for the default rule using a flow hash algorithm. Based on the protocol, source internet protocol address (Source IP Address), source port, destination internet protocol address (Destination IP Address), destination port and transmission control protocol sequence number, select the virtual machine. The transmission control protocol (TCP) connections from a client have different source ports and sequence numbers and can be routed to a different virtual machine. Each individual transmission control protocol connection is routed to a single virtual machine. # VIII. Load Balancing Parameters Focused IX. # Implementational Details Technologies used in this tool are (i). Java programming language, (ii) Java Swing, (iii)Cloud sim (iv) Simjava, (v)Operating system are Windows Xpor Windows7, (vi) IDE: Eclipse, (vii) JDK 1.8 and above. It is aimed to investigate and simulate the large scale internet applications in the cloud environment. Most of the internet applications depend upon many parameters and most of the time the values for those parameters need to be assumed. Therefore it is important to change those parameters and repeat the simulations. Round Robin Algorithm: Round robin is the easiest algorithm used for the logic division of time and nodes. In this algorithm, the overall time is partitioned into the number of segments, and each node in the system is allocated a with a particular time segment or time period. Inputs: Virtual machine image size 10000 (Image size is operating system, application, data to be installed on multiple virtual machines-Usually it is tested for security, reliability and has the best tested configuration), Virtual machine memory -1024 Mb, Virtual machine bandwidth -1000, Virtual machine operating system-Linux, Virtual machine manager -Xen, RAM -204800 Mb, Storage Demerits of round robin algorithm: (i) Clients have to wait in the waiting queue un till and un less the suitable virtual machine is available (ii) The additional load on the scheduler to decide the size of quantum. a) Throttled algorithm and its features (i). It is a static algorithm or static scheduling (iii).Throttled virtual machine load balancer keeps a list of virtual machines and their status whether they are busy or idling when the user cloudlet request comes assigns to a proper virtual machine and lets the work to be done. # Inputs: In case of inputs, same virtual machine configuration and data centre configurations need to give in order to get proper evaluation of throttled algorithm. # b) Results obtained on Throttled algorithm # Combination of round robin and throttled algorithm: As since cloud computing needs heterogeneous nature of virtual machines, throttled algorithm does not support. Round robin algorithm can be combined with throttled algorithm in a hybrid approach. Inputs: In case of inputs, same virtual machine configuration and data centre configurations need to give in order to get proper evaluation of RTH algorithm. # Results obtained from the execution of RTH algorithm: RTH algorithm results are as given below in Table 3. The values stated here are taken on average and measured in nano seconds. Advantages of RTH Algorithm: A slight improvement is observed in makespan waiting time and burst time. Problem statement: It is observed that it requires all virtual machines configuration to be the same then only it is showing better performance. In other words, updating the index table is getting delayed in providing the response to the arrived requests. ESCE algorithm: It is based on the spread spectrum method. In this method, the load balancer monitors the scheduled jobs. Equally Spread Current Execution load balancer puts all the tasks in the job pool and assigns them to the virtual machine. As the load balancer monitors the Scheduled Jobs, balancer keeps track of job's which are in the queue frequently. Inputs: In case of inputs, same virtual machine configuration and data centre configurations need to give in order to get proper evaluation of equally spread current execution algorithm. Problem statement: It is observed that this algorithm is slow to obtain an accurate solution and it requires new fitness tasks on the new algorithm parameters to improve performance. Therefore there is a need for improvement. Need for the improvement: As since the artificial bee colony optimization algorithm requires improvements. As per the existing real time scenario there is a need for the algorithm with the following characteristics (i) The algorithm needs to work efficiently in the distributed environment (ii) The algorithm needs to be a virtual machine friendly, (iii) The algorithm should make the virtual machine to work efficiently under the heavy loads and (iv) The algorithm should assure that it is capable to provide services in the peak demand hours sufficiently. # Combination of RTEH and ABCO Algorithms: The ABCO is a novel and a heuristic algorithm. This algorithm is motivated by honey bee intelligent foraging behaviour. It is used for a searching process to solve the real time parameters optimization problems. The Disadvantages of this algorithm is, it is slow in getting the better result. Many numbers of objective function evaluations are required. There is a possibility of losing relevant information when the function is being optimized. To improve RTEH algorithm, ABCO algorithm can be combined using the hybrid approach. This algorithm can be called as RTEAH algorithm. i. # RTEAH Algorithm Simulation The research is focused on load balancing in hybrid algorithmic approach in cloud computing environment. ii. Efforts are done to make betterment of the performance of scheduling in the cloud environment. iii. Research study is focused on load balancing in hybrid algorithmic approach in cloud computing environment. 2ThrottledMakespan300.38Waiting time0.03Burst time0.25Demerits of throttled algorithm: It works properly only if all virtual machines in data centre have the same hardware configuration. Therefore there is a need to improvement. Hybrid approach can be used to improve this algorithm. Modifications in the algorithm would bring flexibility. 3Makespan300.34Waiting time0.02Burst time0.24RTH algorithm comparison with round robin and throttled algorithm: RTH algorithm comparison with round robin and Throttled algorithms results are as given below in 4RRThrottledRTHMakespan300.37300.38300.34Waiting time0.040.030.02Burst time0.250.250.24 1 4Improved Hybrid Algorithm Approach Based Load Balancing Technique in Cloud Computing 2 5Algorithm Simulation ResultESCEMakespan300.38Waiting time0.03Burst time0.26 6Execution(ESCE) Algorithm ComparisonRRThrottledRTHESCEMakespan 300.37300.38300.34300.38Waiting time0.040.030.020.03Burst time0.250.250.240.26 7RTEHMakespan300.34Waiting time0.02Burst time0.24RTEH Algorithm in comparison with earlier Algorithms:RTEH Algorithm in comparison with earlier algorithmtable are look like as given below in Table 7. The valuesstated here are taken on average and measured in nanoseconds.Table 7: RTEH Algorithm in comparison with EarlierAlgorithmsRRThrottled RTHESCERTEHMakes pan300.37300.38 300.34 300.38295.35Waiting time0.040.030.020.030.02Burst time0.250.250.240.260.24Advantages of new RTEH Algorithm: A slightimprovement is observed in makespan waiting time andburst time. 8Optimization (ABCO) AlgorithmABCOMakespan300.34Waiting time0.02Burst time0.24 10Load balancing in RTEAH AlgorithmRTEAHMakespan292.46Waiting time0.01Burst time0.23RTEAH Algorithm in comparison with otherAlgorithmsRTEAH Algorithm can be compared with otherprevious algorithms as given below in Table.10Comparison of RR, Throttled, RTH, ESCE, RTEHand RTEAH algorithms for the makespan in the form of agraph1. 9295 300 305300.37300.38300.34300.38295.35300.34292.46 Makespan285 290MakespanRRThrottledRTHESCERTEHABCRTEAH * Load Balancing Techniques: Need, Objectives and Major Challenges in Cloud Computing -A Systematic Review NitinKumar Mishra NishcholMishra International Journal of Computer Applications 131 18 December 2015 * Current Scenario in Architect and Applications of Cloud DoddiniProbhuling L International Journal of Advanced Computer and Mathematical Sciences 2230-9624. Vol4 Issue3, 2013 * A Cloud Computing Architecture Framework for Scalable RFID ChetnaDabas Member JPIaeng Gupta IMECS 2010 Proceedings of the International Multi Conference of Engineers and Computer Scientists the International Multi Conference of Engineers and Computer ScientistsHongkong 2010. March 17-19, 2010 1 * A General and Practical Datacenter Selection Framework for Cloud Services HXu BLi 10.1109/cloud.2012.16 IEEE Fifth International Conference on Cloud Computing 2012. 2012 * Load Balanced Min-Min Algorithm for Static Meta-Task Scheduling in Grid Computing TKokilavani DrD IGeorge Amalarethinam International Journal of Computer Applications 20 2 April 2011 * Calculating the minimum bounds of energy consumption for cloud networks JensBuysse KonstantinosGeorgakilas * A novel energy-efficient resource allocation algorithm based on immune clonal optimization for green cloud computing WannengShu WeiWang YunjiWang SPRINGER -EURASIP Journal on Wireless Communications and Networking 2014 * Response Time Minimization Task Scheduling Algorithm MHemamalini MVSrinath Phd International Journal of Computer Applications 145 1 July 2016 * Comparative Analysis of Load Balancing Algorithms in Cloud Computing Ms Nitika Ms MrShaveta Raj International Journal of Advanced Research in Computer Engineering & Technology 1 3 May 2012 * Improved Hybrid DLBS Artificial Bee Colony Optimization Algorithm based on Parallel Computing Environment BhuvneshPathania AbhilashSharma International Journal of Computer Applications 164 3 April 2017 * Genetic Algorithm and Ant Colony Algorithm Based Energy-Efficient Task Scheduling JianfengZhao HongzeQiu Third International Conference on Information Science and Technology Yangzhou, Jiangsu, China IEEE March 23-25. 2013 * Pareto frontier for job execution and data transfer time in hybrid clouds JavidTaheri A AlbertYZomaya A HowardJaySiegel B Zahirtari Future Generation Computer Systems 37 2014. 10 July 2013. October 2013 Accepted 3 December 2013 Available online 18 December 2013 Elsevier Received in revised form 14 * Ant Colony Optimization for Mapping, Scheduling and Placing in Reconfigurable Systems FabrizioFerrandi PierLucaLanzi AntoninoTumeo 2013 * NASA/ESA Conference on Adaptive Hardware and Systems (AHS-2013) 2013 * A Hybrid Scheduling Algorithm to improve the Performance using Ant Colony Optimization and Cuckoo Search Algorithm with K-means++ in a Virtualized Environment APNirmala .SDr Veni International Journal of Advanced Research in Computer Science 7 4 July-August 2016 * Bi-objective HWDO Algorithm for Optimizing Makespan and Reliability of Workflow Scheduling in Cloud Systems PoonamSingh MaitreyeeDutta NaveenAggarwal Pages: 1-10. 16 Conference Paper December2017. March 2014 89 An Analysis of Task Scheduling in Cloud Computing using Evolutionary and Swarm-based Algorithms * Issues and Challenges of Load Balancing Algorithm in Cloud Computing Environment MahfoozAlam Zaki AhmadKhan DOI: 10. 17485 2017/v10i25/105688, July 2017 10 * Load Balancing Techniques in Cloud Computing: A Study BDeepak SShashikala KRadhika International Conference on Information and Communication Technologies (ICICT-2014) * A Methodological Survey on Load Balancing Techniques in Cloud Computing P MohamedShameem RShaji International Journal of Engineering and Technology (IJET) 5 5 Oct-Nov 2013