How to do a Comparative Study of Load Balancing Algorithms in Cloud Computing

In recent years, Cloud Computing has been an emerging computing model in the IT industry. 

 Load Balancing is major concern in the cloud virtual environment Now a day’s many clients all over the world are demanding for various services, Cloud computing provide services by dynamically configure its servers and these servers may be present physically or virtually in the computing environment.

Many algorithms have developed for allocating client's requests to available remote nodes. Efficient load balancing ensures efficient resource utilization of resources to customers on demand basis and enhanced the overall performance of the Cloud.

 Here  a brief discussion about the  existing load balancing techniques in cloud computing and comparison is based on various parameters like data processing time and response time etc. 

Basic comparison between existing Optimal and Throttled scheduling algorithms etc.

The algorithms those are discussed below are previously implemented for load balancing in cloud computing using Cloudanalst tool or Cloudsim tool. 

So here we are discussing the comparative study of them and various environments are defined and compared with each other 

                                                                               TABLE I.

COMPARISION TABLE OF EXISTING LOAD BALANCING ALGORITHMS

	Algorithm	Static environment	Dynamic environment	Distributed balancing
	Round- robin	YES	NO	NO
	Active Clustering	NO	YES	YES
OLB		YES	NO	NO
	CLBDM	YES	NO	NO
	Biased Random Sampling	NO	YES	YES
	MaxMin	YES	NO	NO
	MinMin	YES	NO	NO
	Ant colony	NO	YES	YES
	Stochastic Hill Climbing	NO	YES	NO

                                                                      TABLE II.

PROS AND CONS OF LOAD BALANCING ALGORITHMS

Algorithm	Pros	Cons
Round Robin	Priority basis Sequence processing	Increase waiting time
Throttled Load balancing	Allocate job for suitable VM	The finding is difficult
Active Clustering	Same nodes are grouped Easy to process	Heterogeneous nodes are not counted.
Dual Direction downloading from FTP servers (DDFTP)	Reduce network overhead Reliable to download files	Requires high storage in all nodes
Load balancing Min-Min (LBMM)	Reliable to assign the task	Slower than other algorithms
Equally Spread Current Execution Algorithm	Allocate process based on job waiting	If the job asks for execution, then only allocate process otherwise it did not allocate the process for that occupation.