Task scheduling on heterogeneous devices in parallel pervasive systems (P 2 S)

Abstract

Parallel Pervasive Systems (P 2 S) comprise an ad hoc network of pervasive devices such as cell phones, handheld computers, laptops, sensors and other devices that essentially form a parallel system. Most of the current work in pervasive computing and mobile adhoc networks exploit resources on remote devices to execute compute intensive tasks. In this paper, we present a distributed task scheduling algorithm called cluster based scheduling algorithm (CBS) for parallel processing of task graphs in pervasive environments. We reduce the communication overhead by considering the devices to be grouped into logical clusters. CBS does not require the task scheduler device to have knowledge of all characteristics of each device in the environment. The proposed scheme allows usage of multiple task scheduling algorithms. Simulation results demonstrate time and energy efficient scheduling of tasks in heterogeneous environments. © 2008 Springer Berlin Heidelberg.