Download as pptx, pdf, or txt
Download as pptx, pdf, or txt
You are on page 1of 22

A hybrid elephant herding optimization and

harmony search algorithm for


potential load balancing in
cloud environments

1 Under the guidance of Presented by


Mr Syed Muqthadar Ali Prakash Thirandasu
Senior Assistant Professor 22B81DA902
CSE Dept. CSE Dept
2 CONTENTS
• Abstract
• Introduction
• Literature Survey
• Existing Method
• Motivation
• Problem Statement
• Objectives
• Proposed Method
• Experimental results
• Conclusion
• Future Scope
• References
3 Abstract
 Load balancing becomes a challenge when there is a large influx of new
IoT user requests connected to specific fog nodes.
 An efficient load balancing strategy is required.
 Hybrid Elephant Herding Optimization and Harmony Search Algorithm
for Potential Load Balancing in Cloud Environments (HEHO-HSA-PLB-
CE) is proposed.
 To reduce task waiting time, improve load balancing rates, minimize
scheduling time, decrease delay.
4 Introduction

 Ability to solve large-scale problems and accommodate the explosive


growth of internet usage.
 It provides software applications and hardware resources to cloud users
over the internet.
 To share various resources such as software, servers, applications,
services, storage, networks.
 Effective management of cloud resources is crucial to sustain the scalable
and efficient nature of cloud computing.
5

LITERATURE SURVEY
SNo Title Author Journal Name & Year Methodology Adopted Key Findings Gaps

1 A novel resource discovery mechanism Moorthy R. S Proceedings of the International Conference Sine Cosine Optimization Algorithm Resource Management Importance, Scalability, Real-World Data and Workloads,
using sine cosine optimization algorithm Pabitha P on Intelligent Computing and Control Systems Novel Resource Discovery Mechanism, Comparison with Other State-of-the-Art
in cloud (ICICCS), Madurai, India, IEEE (2020) Comparison of Clustering Algorithms, Algorithms, Algorithm Parameter Tuning,
Dynamic Resource Management
Comparison Metrics, Superiority of
Proposed Approach

2 Load balancing in cloud computing – A Shahbaz Afzal Journal of Cloud Computing: Advances, Round Robin Algorithm Load balancing techniques, highlighting Lack of comprehensive classification ,
hierarchical taxonomical classification G. Kavitha Systems and Applications (2019) Weighted Round Robin Algorithm their advantages, limitations, and Inadequate consideration of causal factors,
challenges Advantages and limitations of existing
methods, Addressing crucial challenges, New
insights and future directions

3 Load balancing in cloud computing Abderraziq John Wiley & Sons, Ltd (2019) Adaptive Starvation Threshold Load Response time, server utilization, Comparative evaluation, Scalability
environments based on adaptive starvation Semmoud Mourad Balancing algorithm migration costs, and system stability. considerations, Real-world
threshold Hakem Badr implementation, Robustness and fault
Benmammar Jean-
tolerance, Consideration of additional
Claude Charr
metrics

4 Implementation of novel load balancing Joshi N International Conference on Computer Particle swarm optimization Stimulated Adoption of Cloud Real-World Validation, Scalability,
technique in cloud computing Kotecha K Communication and Informatics (2018) algorithm Computing, Resource Management and Performance under Heterogeneous
environment Choksi D Load Balancing Challenges, Need for Workloads, Comparison with State-of-
Pandya S
Efficient Load Balancing Techniques, the-Art, Security and Fault Tolerance
Novel Load Balancing Technique,
Evaluation and Comparison, Disaster
Management Capability.
Existing Method
6

Existing load balancing techniques are commonly employed in cloud environments,


including:
1. Round Robin
2. Weighted Round Robin
3. Least Connection
4. Throttling
5. Dynamic Load Balancing
While the aforementioned techniques are established, the hybrid algorithm
combining EHO and HS could offer a novel approach for load balancing in cloud
environments. Its potential lies in leveraging the unique characteristics and
capabilities of both algorithms to adaptively allocate workloads, optimize resource
utilization, and achieve better load balancing performance.
7
Motivation
 Efficient load balancing in cloud environments can lead to several benefits,
including:
1. Enhanced Performance.
2. Resource Optimization.
3. Scalability and Elasticity.
4. Quality of Service (QoS) Improvement.
The motivation for using this hybrid algorithm lies in its potential to improve load
balancing in cloud environments. By developing and applying a hybrid algorithm
that combines the strengths of EHO and HS, the motivation is to address the
challenges of load balancing in cloud environments and optimize resource
allocation. This can lead to improved performance, resource utilization,
scalability, and QoS, ultimately benefiting both cloud service providers and users.
8 Problem Statement

 To implement HEHO-HSA-PLB-CE in cloud environments and


employed for various evaluation metrics, including response time, load
balance rate, scheduling time, delay.

 To assess the performance and effectiveness of the proposed HEHO-


HSA-PLB-CE approach.
Objectives
9

 The HEHO algorithm and HSA are employed to optimize the allocation of
virtual machines (VMs)

 To enhance the strategy for selecting physical machines.

 To overcome challenges such as premature convergence and getting


trapped in local optima.
10 Proposed Method

Block diagram for load balance in cloud computing environment


Proposed Method Cont..
11

 Initialization
 Load Monitoring
 Elephant Herding Optimization (EHO) Component
 Harmony Search (HS) Component
 Hybridization and Decision Making
 Load Redistribution
 Performance Evaluation and Iteration
12 key functionalities of HEHO-HSA-LB

 Categorizing VMs: Categorizing VMs into under-loaded and over-


loaded groups. To balance the workloads across the system.
 Energy minimization: To reduce overall incurred costs. By optimizing
energy usage, the data centre can operate more efficiently.
 Utilization of VMs: The number of under-utilized VMs in the data
centre and transitions their state from active to sleep are
determined. To achieve better resource utilization and reduces
unnecessary energy consumption.
13
Experimental results
14
Performance Analysis
15
Performance Analysis

Scheduling Time
16 Inference from results

 By utilizing the hybrid optimization algorithm and incorporating an


enhanced physical machine selection strategy.
 To improve load balancing in cloud computing environments.
 The evaluation results obtained through the simulation provide insights
into the performance of the proposed approach.
 Potential benefits in terms of response time, load balancing, scheduling,
delay reduction.
17 Supporting Factors

 Virtualization of cloud properties.


 Cloud service providers (CSPs) offer services to users on a rental basis.
 Managing the available virtual cloud resources is a complex issue.
 Aspect of load balancing and maintaining system performance.
 Balance between user satisfaction and financial benefits.
 Agreements with cloud users, managing service level agreements (SLAs),
and ensuring equal distribution of tasks among virtual machines (VMs) or
physical hosts.
18 Conclusion
 HEHO-HSA-PLB-CE to address the challenge of load
balancing in cloud computing environments.

 Reduced task waiting time, improved load balancing rates,


minimized scheduling time, decreased delay.

 Achieved load balancing and resource utilization


enhancement.

 The evaluation metrics indicate improvements in response


time, load balancing, scheduling, delay reduction.
19 Conclusion, cont..

 Balancing workloads among CE hosts.


 Bandwidth, available memory speed, and processor speed.
 Maintaining service quality.
 Optimizing resource utilization.
 Improving system performance.
 Effective resource utilization, and enhanced user experience.
20 Future Scope

The proposed HEHO-HSA-PLB-CE approach presents a significant advancement


in addressing the load balancing challenge in cloud computing environments.
However, there are several areas for future exploration and improvement:
 Scalability Testing
 Real-world Deployment
 Dynamic Workloads
 Comparison with Other Techniques
 Security and Privacy Considerations
 Adaptive Parameters
 Hybrid Approaches
21 References

1. Shafiq D. A., Jhanjhi N. Z., Abdullah A., Alzain M. A., A load balancing algorithm for the
data centers to optimize cloud computing applications, IEEE Access, 9:41731– 41744, 2021.
2. Zhang W. Z., Elgendy I. A., Hammad M., Iliyasu A. M., Du X., Guizani M., Abd El-Latif A.
A., Secure and optimized load balancing for multitier iot and edge-cloud computing systems,
IEEE Internet Things J. 8:8119–8132, 2020.
3. Benkhelifa E., Hani A. B., Welsh T., Mthunzi S., Guegan C. G., Virtual environments testing
as a cloud service: A methodology for protecting and securing virtual infrastructures, IEEE
Access 7:108660–108676, 2019.
4. Son J., Buyya R., Priority-aware VM allocation and network bandwidth provisioning in
software-defined networking (SDN)-enabled clouds, IEEE Trans. Sust. Comput. 4:17–28,
2019.
5. Li W., Cao J., Hu K., Xu J., Buyya R., A trust-based agent learning model for service
composition in mobile cloud computing environments, IEEE Access 7:34207–34226, 2019.
6. Hota A., Mohapatra S., Mohanty S., Computational Intelligence in Data Mining, Advances in
Intelligent Systems and Computing, Vol. 711, Springer, Singapore, 2019, pp. 99–110.
22

THANK YOU

You might also like