Service Assignment Problem: Exploiting Group Context Information in Service Discovery for Multi-user Multi-service Ubiquitous Computing Environments

We consider in this paper the problem of exploiting group context, context information related to interference and conflict among actions of a group of users, in ubiquitous computing environments. We show that integrating group context to service discovery can be accomplished by solving service assignment problem, a problem that shows requirements of assigning services to requesting users appropriately when there are multiple users requesting for multiple candidate services. A service assignment scheme, named QASA, is proposed as a solution to the problem. It leverages query aggregation to handle the asynchronous nature of query arrival patterns and uses an assignment algorithm to allocate services to users. Our experiments show that QASA improves user satisfaction, service utilization and quality of query results in high load conditions without harming those metrics in low and moderate load conditions. QASA also enhances system performance by reducing redundant accesses to service discovery function.