Call For Paper Volume:4 Issue:8 Aug'2017 |

Adaptive Algorithm and a Genetic Algorithm for Minimizing Cloud Task Length with Prediction Error

Publication Date : 01/10/2015



Author(s) :

SHARFAN P.S , Shahina K.K.


Volume/Issue :
Volume 2
,
Issue 9
(10 - 2015)



Abstract :

In a cloud data centre model, server has to process user request efficiently. So task execution length optimization required. But optimizing the task length is difficult due to the involvement of constraints like user payment and divisible resource demand. In Adaptive prediction (AP) method, a local optimal allocation algorithm (LOAA) and a Dynamic optimal divisible resource allocation methods (DODRA) are used. An error prediction mechanism is implemented here through which optimal node allocation done. But still the system is inefficient due to requirement of time and it benefit to the clients only. In my proposed system a genetic partitioning (GP) method is implemented. In GP, after partitioning the cloud system, nodes are allotted for different partitions and job is chosen based on the partition in which node placed. Then by applying genetic algorithm based on these partitions, every task can be optimally allocated to the hosts. Here by comparing both existing and my proposed systems through CloudSim simulator we can analyze that genetic partitioning is efficient than AP method


No. of Downloads :

2


Indexing

Web Design MymensinghPremium WordPress ThemesWeb Development

Adaptive Algorithm and a Genetic Algorithm for Minimizing Cloud Task Length with Prediction Error

September 30, 2015