A Task Allocation Algorithm with Weighted Average Velocity Based on Online Active Period

Abstract

In some complex scientific calculation, the resources of the calculation are very large. To a certain extent, the improvement of the computer level has met the needs of many computing, but a lot of more complex calculation cannot still be effectively resolved. Volunteer computing is a computational method that divides the complexity of computing tasks into simple subtasks, and collects the results of volunteer computing resources to solve the subtasks. In this calculation process, the task assignment module is an extremely important part of the whole computing platform. Many of the existing task allocation algorithms (TAA) are used to group by the similar conditions of the volunteer computer. TAA used in this work grouped by the computers with similar online active period, and the computation efficiency is improved by using the weighted average velocity as a parameter. The experimental results showed that TAA with the weighted average velocity based on online active period can effectively improve the performance of the volunteer computing platform.

Keywords: Volunteer computing; Task allocation algorithm; Weighted average velocity; Online active period