An Adaptive and Distance-based Resource Allocation Scheme for Interdependent Tasks in Mobile Ad Hoc Computational Grids

Abstract

Two key components contribute to task completion time: execution cost and communication cost. The communication cost is induced by data transfers between tasks residing on separate nodes. The communication is always expensive and unreliable in mobile ad hoc Grids and therefore plays a critical role in application performance. To reduce communication cost, interdependent tasks are allocated to nodes located close to one another. However, once the tasks have been allocated, nodes can move within a Grid. The movement of nodes within a Grid may result in multi-hop communication between nodes executing dependent tasks. In order to deal with node mobility within a Grid, an effective resource allocation scheme is required, but the design of such a scheme for mobile ad hoc computational Grids is challenging due to the constrained communication environment, node mobility, and infrastructure-less network environment. In this paper, we have developed an adaptive and distance-based resource allocation scheme which takes into account the characteristics of an application and nodes and applies migration heuristics to address the local node mobility problem. The scheme is validated in a simulated environment using various workloads and parameters.