An Optimized Mobile Cloud Computational Offloading Framework using K-Means Algorithm

Abstract

Offloading the execution of heavy computational modules from mobile devices to Mobile Cloud Computing (MCC) is inevitable in today's era as it mainly focuses in consuming less battery power and execution time. But, the problem incurred with identifying the most optimal cloud device to map each module still remains a challenge in cloud computing environment. In this paper, a novel MCC offloading framework is proposed to fasten the allocation and execution of high computational modules that runs in the mobile device, effectively on the cloud. The framework employs K-Means clustering algorithm to group the nearest cloud virtual machines that best suits for executing modules of software running in the mobile. The objective of the paper is to maximize the energy savings by extending the battery life and execution speed of mobile device when executing heavy computational modules. The optimal selection of cloud device is attained by grouping the requirements of each module with the nearest cloud devices offering the same requirements using K-Means Algorithm. The proposed framework is compared with the existing mobile computation offloading frameworks with respect to energy saving, execution time and energy consumption. The results show that the proposed work executes the modules of computationally intensive modules in minimum time span with maximized energy savings than the existing frameworks.

Computer User Behavior Anomaly Detection Based on K-Means Algorithm

“…erefore, by establishing behavioral patterns, discover inconsistencies between patterns and implement anomaly detection of user behavior [8]. Veeraiyan et al believe that currently, the shortcomings of user behavior anomaly detection methods are: insufficient ability to automatically process large amounts of data [9]. Seta and Hartomo indicated that a user behavior anomaly detection algorithm based on pattern mining can automatically process massive user behavior audit data; in order to a certain extent, the detection accuracy can be improved [10].…”