Distributed deadlock handling for resource allocation in smart spaces

Abstract

In a ubiquitous system, there are several interacting computational objects which use each others' resources. As the number of resources and their consumers grow in such systems, the delay that the consumers experience for obtaining control over resources increases with an existing rule-based resource allocation technique. Distributing the resource allocation, however, complicates the nature of deadlocks that may arise and requires more sophisticated techniques as compared to a setup with central control. The goal of this paper is to generalize the current resource allocation method to a distributed setting and, in particular, to propose an approach for handling deadlocks in this case. © Springer-Verlag Berlin Heidelberg 2011.