Optimizing Hadoop MapReduce for Resource-Constrained Mobile Cloud Environments

Abstract

This research investigates the adaptation and optimization of Hadoop MapReduce for mobile cloud environments characterized by their heterogeneous resources and dynamic nature. The objective is to develop a resource-aware scheduling and task allocation strategy that improves performance and efficiency while minimizing energy consumption. We propose a novel algorithm that considers both processing capacity and network bandwidth constraints, enhancing the scalability and reliability of MapReduce in mobile clouds. Our results demonstrate significant improvements in job completion time and energy efficiency compared to existing approaches.

Introduction

Mobile cloud computing leverages the collective processing power of mobile devices to execute computationally intensive tasks. However, utilizing the inherent heterogeneity and limited resources (battery, bandwidth, processing power) of mobile devices presents significant challenges. Hadoop MapReduce, a widely used framework for distributed data processing, is not inherently optimized for such dynamic and constrained environments. Existing approaches struggle to efficiently manage resources, leading to prolonged job execution times and increased energy consumption. This research aims to address these limitations by developing a tailored MapReduce framework for mobile clouds.

Objectives

• Develop a resource-aware scheduling algorithm for Hadoop MapReduce in mobile cloud environments.
• Optimize task allocation to minimize job completion time and energy consumption.
• Evaluate the performance of the proposed system through simulations and/or experiments.