Knapsack problems: algorithms and computer implementations
Knapsack problems: algorithms and computer implementations
Physical Prototyping with Smart-Its
IEEE Pervasive Computing
Nile-PDT: a phenomenon detection and tracking framework for data stream management systems
VLDB '05 Proceedings of the 31st international conference on Very large data bases
Home network: road to ubiquitous world
VLDB '06 Proceedings of the 32nd international conference on Very large data bases
Pervasive Service Composition in the Home Network
AINA '07 Proceedings of the 21st International Conference on Advanced Networking and Applications
A novel memory management scheme for residential gateways
Information Systems Frontiers
Adaptive quality of service management for next generation residential gateways
MMNS'06 Proceedings of the 9th IFIP/IEEE international conference on Management of Multimedia and Mobile Networks and Services
The Open Services Gateway Initiative: an introductory overview
IEEE Communications Magazine
Finance e-learning and simulation toward the cloud service environment
International Journal of Internet Protocol Technology
An efficient privacy preserving Pub-Sub system for ubiquitous computing
International Journal of Ad Hoc and Ubiquitous Computing
| Hi-index | 0.00 |
Applications in ubiquitous computing environments are expected to have different priorities, memory and real-time requirements. Given a home gateway with a set of running services, we need to terminate some of the existing services to allow a more important service to start. This paper presents heuristics for service replacement that try to terminate the least number of low priority services. Moreover, we develop an integer linear programming (ILP) formulation to use it as a benchmark, against which the quality of solutions obtained by our proposed algorithms is assessed. Our simulation results indicate that the proposed algorithms outperform the traditional memory management techniques in terms of the quality of the obtained solutions. In fact, the performance of the proposed heuristics is close to that of the lower bound achieved by the ILP solution, while the execution times of the proposed algorithms were close to that of the traditional memory management techniques.