JavaScript is disabled for your browser. Some features of this site may not work without it.
Please be advised that the site will be down for maintenance on Sunday, September 1, 2024, from 08:00 to 18:00, and again on Monday, September 2, 2024, from 08:00 to 09:00. We apologize for any inconvenience this may cause.
Ant-inspired strategies for opportunistic load balancing in the distributed computation of solutions to embarrassingly parallel problems
Computational science is a practice that requires a large amount of computing time. One means of providing the required computing time is to construct a distributed computing system that utilises the ordinary desktop computers found within an organisation. However, when the constituent computers do not all perform computations at the same speed, the overall completion time of a project involving the execution of tasks by all of the computers in the system becomes dependent on the performance of the slowest computer in the network. This study proposes two ant-inspired algorithms for dynamic task allocation that aim to overcome the aforementioned dependency. A procedure for tuning the free parameters of the algorithms is specified and the algorithms are evaluated for their viability in terms of their effect on the overall completion time of tasks as well as their usage of bandwidth in the network.
Description:
Dissertation (MSc(Computer Science))--University of Pretoria, 2016.