Network Aware Resource Allocation Scheme for Mobile Ad Hoc Computational Grid

Abstract

Mobile ad hoc computational Grids are characterized by unreliable and dynamic communication environment. The bandwidth at different network portions varies over the time and different nodes often experience different connection quality at the same time due to traffic load and constrained communication environment. Therefore, an effective and robust resource allocation scheme should consider connection quality between nodes in addition to other factors such as transmission power, processing speed and task size. In this paper, two resource allocation schemes are described with their pros and cons. The first resource scheme exploits transmission power control mechanism in order to improve energy efficiency and network capacity whereas the second resource allocation scheme takes into account the connection quality and traffic load in order to reduce data transfer time. The relationship between data transfer time and transmission energy consumption has also been investigated in detail. The schemes are validated in a simulated environment using various workloads and parameters.A distributed computing infrastructure which allows mobile robots to share computing resources on demand without pre-existing network infrastructure is known as a mobile ad hoc computational Grid [4]. Mobile ad hoc computational Grid is an integration of computational Grid and mobile ad hoc network. Computational Grid is a software infrastructure that allows distributed computing devices to share computing resources to solve computationally-intensive problems. Whereas mobile ad hoc network is a wireless network of mobile devices that communicate with each other without any pre-existing network infrastructure.Resource allocation is a central task in any parallel and distributed computing system. Compared to traditional infrastructure-based distributed systems such as Grids and Clusters, mobile ad hoc computational Grids are characterized by shared and unreliable communication medium, low bandwidth, high latency, limited battery power, node mobility and infrastructure-less network environment. Therefore, resource allocation schemes proposed for traditional distributed systems cannot be applied to mobile ad hoc computational Grids.Previously we proposed a power-based resource allocation scheme [4] to improve energy efficiency and network capacity. The basic idea is to exploit transmission power control mechanism and allocate data-intensive dependent tasks to nodes accessible at minimum transmission power. Since energy consumption is proportional to transmission power, communication at minimum required transmission power reduced energy consumed in transmission of data.Since mobile ad hoc computational Grids are characterized by unreliable and dynamic communication environment, bandwidth at different network portions varies over the time and different nodes often experience different connection quality at the same time due to traffic load and constrained communication environment. Therefore, an effective and robust resource allocation scheme should co...