A Multi-Task Oriented Framework for Mobile Computation Offloading

Junyu Lu,Qiang Li,Bing Guo,Jie Li,Yan Shen,Gongliang Li,Hong Su

DOI: https://doi.org/10.1109/tcc.2019.2952346

IF: 5.697

2022-01-01

IEEE Transactions on Cloud Computing

Abstract:Computation offloading has become popular in recent years as it is an effective way to reduce the energy consumption and enhance the performance of smartphones. To deal with the heterogeneous architectures between the smartphone and the server, and to simplify deployment of the server, we propose and implement a lightweight offloading framework which supports offloading of compute-intensive tasks and deploying the server efficiently. Based on this framework, generic and developer-customized offloading services could be provided for different third-party applications. Furthermore, we design a multi-task offloading tactic for the framework to deal with intensive offloading requests from various mobile devices. When receiving an offloading request, the master node in server-side determines whether this task should be offloaded or not and which VM should handle this task, so that the overall execution time and energy consumption are optimized. We implement this framework and evaluate it by comparing the execution time, energy consumption and CPU utilization rate among three execution modes with three applications. We also conduct experiments of the multi-task offloading tactic in simulation environment. Experimental results indicate that this framework effectively reduces energy consumption and boosts performance for compute-intensive tasks, and the multi-task offloading tactic is valid for intensive offloading requests.

computer science, information systems, theory & methods

What problem does this paper attempt to address?

### Problems Addressed by the Paper

The paper primarily aims to address the issue of resource constraints on smartphones when handling computationally intensive tasks. Specifically, the authors propose a lightweight mobile computation offloading framework to tackle the following challenges:

1. **Heterogeneous Architecture Issue**: There are different hardware architectures (such as ARM and x86) between smartphones and servers, making it very difficult to directly match execution contexts.
2. **Improving Performance and Reducing Energy Consumption**: By offloading computationally intensive tasks to remote servers or personal computers, the performance of smartphones can be improved and energy consumption reduced.
3. **Multi-task Offloading Strategy**: A multi-task offloading strategy is designed to effectively handle a large number of offloading requests from different mobile devices and optimize overall execution time and energy consumption.

### Main Contributions

1. **Compatibility with Heterogeneous Architectures**: The framework can achieve dynamic application context deployment between smartphones and x86 computers.
2. **Three Execution Modes**: Three different execution modes are proposed, assigning different execution methods for different types of tasks.
3. **Multi-task Offloading Algorithm**: An effective multi-task offloading strategy based on the Lyapunov optimization method is designed to handle a large number of offloading requests and optimize overall performance.

### Summary

This paper aims to solve the problem of resource constraints on smartphones when handling computationally intensive tasks through a lightweight computation offloading framework, improving performance and reducing energy consumption. Additionally, a multi-task offloading strategy is proposed to efficiently handle a large number of task requests across multiple devices.